CN115392582B - Crop yield prediction method based on increment fuzzy rough set attribute reduction - Google Patents

Crop yield prediction method based on increment fuzzy rough set attribute reduction Download PDF

Info

Publication number
CN115392582B
CN115392582B CN202211066091.XA CN202211066091A CN115392582B CN 115392582 B CN115392582 B CN 115392582B CN 202211066091 A CN202211066091 A CN 202211066091A CN 115392582 B CN115392582 B CN 115392582B
Authority
CN
China
Prior art keywords
crop
fuzzy
subset
crop growth
crop yield
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202211066091.XA
Other languages
Chinese (zh)
Other versions
CN115392582A (en
Inventor
赵洁
叶文浩
赵艮平
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangdong University of Technology
Original Assignee
Guangdong University of Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guangdong University of Technology filed Critical Guangdong University of Technology
Priority to CN202211066091.XA priority Critical patent/CN115392582B/en
Publication of CN115392582A publication Critical patent/CN115392582A/en
Application granted granted Critical
Publication of CN115392582B publication Critical patent/CN115392582B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/02—Agriculture; Fishing; Forestry; Mining

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Theoretical Computer Science (AREA)
  • Marketing (AREA)
  • General Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • Tourism & Hospitality (AREA)
  • Animal Husbandry (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Agronomy & Crop Science (AREA)
  • Mining & Mineral Resources (AREA)
  • General Health & Medical Sciences (AREA)
  • Marine Sciences & Fisheries (AREA)
  • Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Development Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The application provides a crop yield prediction method based on delta fuzzy rough set attribute reduction. The method comprises the following steps: constructing a decision table; inputting a crop growth record data subset as an original data subset into a fuzzy rough set attribute reduction model; calculating fuzzy similarity of each crop growth record under crop yield index data in the original data subset and weighted consistency degree of each crop growth record to obtain a weighted consistency degree and a qualified crop growth record subset; calculating weighted consistency degree and crop yield index importance of the crop growth record subset of the grid, and performing attribute reduction to obtain a reduced attribute subset; and sequentially inputting the remaining crop growth record data subsets serving as incremental data subsets into a fuzzy rough set attribute reduction model for attribute reduction to obtain a plurality of non-redundant reduction attribute subsets, and training to obtain a crop yield prediction model. The application can improve the prediction precision of crop yield and reduce the technical cost.

Description

Crop yield prediction method based on increment fuzzy rough set attribute reduction
Technical Field
The application relates to the technical field of crop yield prediction, in particular to a crop yield prediction method based on increment fuzzy rough set attribute reduction.
Background
China is a large agricultural country, grain production is a basic guarantee for ensuring the high-speed development of economy, and the change of grain yield directly affects the stable development of national economy. At present, along with the rapid development of remote sensing technology, the remote sensing technology is widely applied to agricultural production, and an effective way is provided for monitoring the growth vigor of crops and predicting the yield of seeds. The existing method for predicting crop yield based on remote sensing technology can directly input all preprocessed multi-source remote sensing data into a machine learning model to analyze crop yield, wherein the number of crop yield analysis indexes can reach hundreds, and redundant indexes can seriously increase the technical cost of data acquisition and the time cost of calculation. Therefore, it is critical to select as few indices as possible for crop yield prediction while maintaining high classification accuracy.
Disclosure of Invention
The embodiment of the application provides a crop yield prediction method based on increment fuzzy rough set attribute reduction, which is used for solving the technical problem of high technical cost of the existing crop yield prediction method.
In a first aspect, an embodiment of the present application provides a crop yield prediction method based on delta fuzzy rough set attribute reduction, including:
step S1: acquiring crop growth record data, crop yield index data and crop yield grade data, constructing a decision table, and setting a fuzzy equivalence relation and a fuzzy similarity threshold;
step S2: dividing the crop growth record data into a plurality of crop growth record data subsets with equal sizes according to the decision table;
step S3: inputting one of a plurality of crop growth record data subsets as an original data subset into a fuzzy rough set attribute reduction model constructed based on a decision table, a fuzzy equivalence relation and a fuzzy similarity threshold;
step S4: calculating fuzzy similarity of each crop growth record in the original data subset under crop yield index data through the fuzzy rough set attribute reduction model, calculating weighted consistency degree of each crop growth record in the original data subset according to the fuzzy similarity, and eliminating crop growth records with unqualified consistency degree in the original data subset to obtain a crop growth record subset with weighted consistency degree and qualified consistency degree;
Step S5: according to the decision table, calculating the weighted consistency degree and the importance of the crop yield indexes of the crop growth record subset of the grid, and carrying out attribute reduction on the crop yield indexes according to the importance of the crop yield indexes to obtain a reduced attribute subset;
step S6: performing redundancy check on the reduced Jian Shuxing subset to obtain a non-redundant reduced attribute subset;
step S7: and sequentially inputting the rest crop growth record data subsets serving as incremental data subsets into the fuzzy rough set attribute reduction model, updating the original data subsets, repeating the steps S4 to S6 by using the updated original data subsets to obtain a plurality of non-redundant reduction attribute subsets, and inputting the non-redundant reduction attribute subsets into the classifier model to train to obtain the crop yield prediction model.
In one embodiment, the obtaining crop growth record data, crop yield index data, crop yield grade data, and constructing a decision table includes:
acquiring crop growth record data, and clustering the crop growth record data according to crop yield to obtain crop yield grade data;
normalizing the crop yield index value of the crop growth record data to obtain crop yield index data;
And constructing a decision table DT= (U, C U D) according to the crop growth record data, the crop yield index data and the crop yield grade data, wherein DT represents the decision table, U represents the crop growth record set, C represents the crop yield index set, and D represents the crop yield grade set.
In one embodiment, the calculating, by the fuzzy rough set attribute reduction model, fuzzy similarity of each crop growth record in the original data subset under the crop yield index data, calculating weighted consistency degree of each crop growth record in the original data subset according to the fuzzy similarity, and removing crop growth records with unqualified weight consistency degree in the original data subset to obtain a subset of crop growth records with weighted consistency degree and qualified weight consistency degree includes:
calculating the fuzzy similarity of each crop growth record in the original data subset under the crop yield index data through the fuzzy rough set attribute reduction model to obtain a fuzzy set;
sorting crop growth records in the fuzzy set according to the size of the fuzzy similarity to obtain a weighted fuzzy set;
calculating the weighted consistency degree of each crop growth record in the original data subset with the crop yield grade under the crop yield index data according to the weighted fuzzy set;
And removing crop growth records with the weight consistency degree failing in the original data subset according to the weight consistency degree to obtain a crop growth record subset with the weight consistency degree failing.
In one embodiment, the calculating, by using the fuzzy rough set attribute reduction model, fuzzy similarity of each crop growth record in the original data subset under the crop yield index data, calculating weighted consistency degree of each crop growth record in the original data subset according to the fuzzy similarity, and removing crop growth records with unqualified weight consistency degree in the original data subset to obtain a subset of crop growth records with weighted consistency degree and qualified weight consistency degree, and further includes:
and calculating the weighted consistency degree and the fuzzy positive area of the crop growth record subset under the crop yield index data according to the fuzzy similarity of each crop growth record in the original data subset under the crop yield index data.
In one embodiment, the calculating the weighted consistency degree and the crop yield index importance of the subset of the crop growth records according to the decision table, and performing attribute reduction on the crop yield index according to the crop yield index importance, to obtain a reduced attribute subset includes:
Initializing an original reduced attribute set;
calculating the fuzzy positive areas of the weighted consistency degree and the grid crop growth record subsets under the original reduced attribute sets according to the fuzzy positive areas of the weighted consistency degree and the grid crop growth record subsets under the crop yield index data, and obtaining updated crop growth record subsets;
calculating the importance of crop yield indexes of the updated crop growth record subset according to the weighted consistency degree and the fuzzy positive area of the crop growth record subset under the original reduced attribute set;
and according to the importance of the crop yield index, carrying out attribute reduction on the crop yield index until a stopping condition is reached, so as to obtain a reduced attribute subset.
In one embodiment, said performing redundancy check on said reduced Jian Shuxing subset to obtain a non-redundant reduced attribute subset comprises:
screening out redundancy attributes from the approximately Jian Shuxing subset according to redundancy check conditions;
and eliminating the redundant attribute from the reduced attribute subset to obtain a non-redundant reduced attribute subset.
In one embodiment, the redundancy check condition is specifically:
and when the fuzzy positive area of the crop growth record subset with the weighted consistency degree and the grid is larger than zero under the crop yield index data, and the importance of the crop yield index after one of the reduced attributes is removed from the about Jian Shuxing subset is equal to the importance of the crop yield index data, judging that the removed reduced attribute is a redundant attribute.
In a second aspect, an embodiment of the present application provides a crop yield prediction apparatus based on delta fuzzy rough set attribute reduction, including:
the decision table construction module is used for executing step S1: acquiring crop growth record data, crop yield index data and crop yield grade data, constructing a decision table, and setting a fuzzy equivalence relation and a fuzzy similarity threshold;
the crop growth record data dividing module is used for executing the step S2: dividing the crop growth record data into a plurality of crop growth record data subsets with equal sizes according to the decision table;
the first input module is used for executing step S3: inputting one of a plurality of crop growth record data subsets as an original data subset into a fuzzy rough set attribute reduction model constructed based on a decision table, a fuzzy equivalence relation and a fuzzy similarity threshold;
the crop growth record rejecting module is used for executing the step S4: calculating fuzzy similarity of each crop growth record in the original data subset under crop yield index data through the fuzzy rough set attribute reduction model, calculating weighted consistency degree of each crop growth record in the original data subset according to the fuzzy similarity, and eliminating crop growth records with unqualified consistency degree in the original data subset to obtain a crop growth record subset with weighted consistency degree and qualified consistency degree;
The attribute reduction module is configured to execute step S5: according to the decision table, calculating the weighted consistency degree and the importance of the crop yield indexes of the crop growth record subset of the grid, and carrying out attribute reduction on the crop yield indexes according to the importance of the crop yield indexes to obtain a reduced attribute subset;
the redundancy check module is used for executing step S6: performing redundancy check on the reduced Jian Shuxing subset to obtain a non-redundant reduced attribute subset;
a loop module, configured to execute step S7: and sequentially inputting the rest crop growth record data subsets serving as incremental data subsets into the fuzzy rough set attribute reduction model, updating the original data subsets, repeating the steps S4 to S6 by using the updated original data subsets to obtain a plurality of non-redundant reduction attribute subsets, and inputting the non-redundant reduction attribute subsets into the classifier model to train to obtain the crop yield prediction model.
In a third aspect, an embodiment of the present application provides an electronic device, including a processor and a memory storing a computer program, where the processor implements the crop yield prediction method based on the reduction of the delta fuzzy rough set attribute according to the first aspect when executing the program.
In a fourth aspect, an embodiment of the present application provides a computer program product, including a computer program, which when executed by a processor implements the crop yield prediction method based on the reduction of the incremental fuzzy rough set attribute according to the first aspect.
According to the crop yield prediction method based on the increment fuzzy rough set attribute reduction, provided by the embodiment of the application, the crop growth record data is divided into a plurality of crop growth record data subsets with equal sizes by constructing the decision table, so that the data volume for attribute reduction each time can be reduced, the technical difficulty is reduced, and the calculation time is shortened; when the attribute is reduced, one crop growth record data subset is used as an original data subset to be input into a fuzzy rough set attribute reduction model for first attribute reduction, the rest crop growth record data subset is used as an incremental data subset to be sequentially input into the fuzzy rough set attribute reduction model, next attribute reduction is performed after the original data subset is updated until the incremental data subset is empty, technical difficulty in the calculation process is reduced, calculation efficiency is improved, more accurate attribute reduction results can be obtained, and accuracy of a crop yield prediction model is improved; redundancy check is carried out on the reduced attribute subset after attribute reduction is carried out each time, and under the condition that the attribute reduction calculation is carried out by increasing the incremental data subsets, the simplified non-redundant reduced attribute subset can be ensured to be obtained, and further long-term effectiveness and reliability of the algorithm under a dynamic environment are ensured.
Drawings
In order to more clearly illustrate the application or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the application, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a crop yield prediction method based on delta fuzzy rough set attribute reduction provided by an embodiment of the application;
FIG. 2 is a schematic structural diagram of a crop yield prediction device based on incremental fuzzy rough set attribute reduction according to an embodiment of the present application;
fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present application more apparent, the technical solutions of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
Fig. 1 is a flow chart of a crop yield prediction method based on delta fuzzy rough set attribute reduction according to an embodiment of the present application. Referring to fig. 1, an embodiment of the present application provides a crop yield prediction method based on delta fuzzy rough set attribute reduction, which may include:
step S1: acquiring crop growth record data U, crop yield index data C and crop yield grade data D, constructing a decision table DT= (U, C U D), and giving a fuzzy equivalence relationAnd a fuzzy similarity threshold α;
step S2: dividing the crop growth record data into a plurality of crop growth record data subsets with equal size according to the decision table (in two cases, because the time complexity of the algorithm is very high, (1) if the existing data set is very large, the large data set needs to be divided into small data sets to be calculated in an incremental mode, so that the problem that the calculation cannot be performed due to insufficient hardware conditions is avoided;
step S3: inputting one of a plurality of crop growth record data subsets as a raw data subset into a decision table DT-based fuzzy equivalence relation And a fuzzy rough set attribute reduction model constructed by a fuzzy similarity threshold alpha;
step S4: calculating the original data subset U through the fuzzy rough set attribute reduction model k Each crop growth is recorded in crop yield index data DT= (U, C U D) C= { a 1 ,a 2 ,...,a m The fuzzy similarity under the condition is ranked, the weighted consistency degree of crop growth records is calculated according to the fuzzy similarity, a consistency degree threshold value eta is set, and the original data subset U is subjected to the weighted consistency degree k Crop growth record set divided into weighted consistency degree inequality (less than eta)And a set of crop growth records weighted for degree of consistency and grid (η or more)>Crop growth record set with disqualified weighted consistency degree>
Step S5: according to the decision table, calculating a crop growth record set with weighted consistency degree and gridThe importance of the crop yield index is reduced according to the importance of the crop yield index, and the optimal crop yield index is selected until the stopping condition is met, so that a reduced attribute subset red' is obtained;
step S6: performing redundancy check on the reduced Jian Shuxing subset to obtain a non-redundant reduced attribute subset red';
Step S7: subset U of the crop growth record data remaining k+1 Sequentially inputting the fuzzy rough set attribute reduction model as an incremental data subset, updating the original data subset, repeating the steps S4 to S6 by using the updated original data subset to obtain a non-redundant reduction attribute subset red', and inputting the non-redundant reduction attribute set into a classifier model to train to obtain a crop yield prediction model for predicting crop yield.
It should be noted that, the execution subject of the crop yield prediction method based on the delta fuzzy rough set attribute reduction provided by the embodiment of the present application may be any network side device, such as a crop yield prediction system.
In step S1, the network side device acquires crop growth record data U, crop yield index data C, and crop yield grade data D, constructs a decision table dt= (U, C U D), and gives a fuzzy equivalence relationAnd a fuzzy similarity threshold α.
It should be noted that, in order to evaluate crop yield according to the existing data, a decision table dt= (U, C U D) may be defined first, where U is a field, is a set of non-empty limited crop growth records, C is a set of non-empty limited crop yield indexes (attributes), and D represents a crop yield level. Determining fuzzy equivalence relation Each condition attribute a.epsilon.C can define a fuzzy relation +.>Whereby arbitrary property subset->Corresponding fuzzy relationSimilarly, attribute set C corresponds to a fuzzy relationship of +.>Let u= { x 1 ,x 2 ,...,x n The real value set of n crop growth records is c= { a 1 ,a 2 ,...,a m The crop yield index set of the crop fuzzy information system mainly comprises a vegetation index, a red edge index, a soil adjustment vegetation index, an enhanced vegetation index, an anti-atmosphere index, a greenness vegetation index, a desertification vulnerability index and a green ratioAnd (5) calculating m crop yield indexes such as vegetation indexes, spectrum indexes and the like. D= { D 1 ,d 2 ,...,d r And r crop yield levels of the decision table. Let U's decision be divided into U/D= { [ x] D :x∈U}，[x] D = { y e U: D (x) =d (y) }, D is the decision class of crop growth record x, [ x ]] D Is defined as the membership ofWherein y ε U. [ x ]] D = { y e U: D (x) =d (y) } represents the set of objects for which x is at the crop yield level (if the decision attribute of the object is the same as x, it belongs to the set, otherwise it does not). Fuzzy equivalence relation->Generating a fuzzy division on U> Is a fuzzy set on U. Arbitrary y ε U, y is toMembership of +.>Given alpha E [0,1 ]]，/>Alpha-grade common set of (2) is +. > Is the alpha-level fuzzy equivalence class for x. Wherein (1)>
In one embodiment, S1 may include:
acquiring crop growth record data, and clustering the crop growth record data according to crop yield to obtain crop yield grade data;
normalizing the crop yield index value of the crop growth record data to obtain crop yield index data;
constructing a decision table DT= (U, C U) D according to the crop growth record data, the crop yield index data and the crop yield grade data, wherein DT represents a fuzzy decision table, U represents a crop growth record set, C represents a crop yield index set, D represents a crop yield grade set, and a fuzzy equivalence relation is givenAnd a fuzzy similarity threshold α.
It should be noted that the acquired data may be multi-source remote sensing data, and the network side device may cluster the crop growth record data according to the crop yield by using a K-means (K-means) clustering algorithm, that is, divide the crop growth record data into r yield levels d= { D according to the crop yield 1 ,d 2 ,...,d r }。
It should be noted that, the network side device may, for each crop yield index a e C, set the value range of crop yield index a to [ a ] - ,a + ]Then the crop yield index value a (x i ) Normalization can be performed by the following formula (1).
Wherein a (x i ) Indicating crop yield index value, a' (x) i ) Represents normalized crop yield index value, a' (x) i )∈[0,1]U represents a crop growth record set, x i Crop growth records are shown. If the data is dynamic flow data, the crop yield index of the new data set is increasedThe threshold value must be in the historical value range [ a ] - ,a + ]And if not, the historical data is re-integrated for calculation.
The network side equipment can obtain real-value data of multi-source remote sensing through normalization, the data information quantity is reserved, the information loss needing discretization is avoided, and the reliability and the dependability of crop estimation by using a small quantity of crop yield index results are ensured.
For convenience of notation, the subsequent a represents the crop yield index after normalization.
In step S2, the network side device divides the crop growth record data into a plurality of equal-sized crop growth record data subsets according to the decision table.
The method is applicable to two cases, and the time complexity of the fuzzy rough set attribute reduction algorithm is very high: (1) If the existing data set is very large, the large data set is required to be divided into small data sets for calculation, and calculation is performed in an incremental mode, so that the problem that calculation cannot be performed due to insufficient hardware conditions is avoided; (2) If the data is streaming data that dynamically increases over time, the new data set is incrementally computed on the results of the original data set. In order to ensure smooth algorithm calculation and improve calculation efficiency, the crop growth record data can be divided into a plurality of crop growth record data subsets according to actual conditions.
It should be noted that, the network side device may divide the crop growth record set U into t copies according to actual situations, so that the crop growth record data is randomly divided into t crop growth record data subsets with equal size, and a crop sample subset sequence u= { U is formed by using the t crop growth record data subsets 1 ,U 2 ,...,U t }. If the crop growth record data volume is too large, a mode of dividing multiple data can be adopted, so that the running cost of a computer can be effectively reduced, and large-scale calculation can be performed.
In step S3, the network side device will input one of the crop growth record data subsets as the original data subset into the decision table DT-based fuzzy equivalence relationAnd a fuzzy rough set attribute reduction model constructed by the fuzzy similarity threshold alpha.
When attribute reduction is carried out subsequently, one crop growth record data subset is used as an original data subset to be input into the fuzzy rough set attribute reduction model for carrying out first attribute reduction, the rest crop growth record data subset is used as an increment data subset to be sequentially input into the fuzzy rough set attribute reduction model, next attribute reduction is carried out after the original data subset is updated until the increment data subset is empty, technical difficulty in a calculation process is reduced, calculation efficiency is improved, a more accurate attribute reduction result can be obtained, and accuracy of a crop yield prediction model is improved.
In step S4, the network device calculates the original data subset U by using the fuzzy rough set attribute reduction model k Each crop growth is recorded in crop yield index data c= { a 1 ,a 2 ,...,a m Fuzzy similarity under } and ordering. Setting a weighted consistency degree threshold eta, calculating the weighted consistency degree of crop growth records according to the fuzzy similarity, and carrying out the primary data subset U according to the weighted consistency degree k Crop growth record set divided into weighted consistency degree inequalityAnd a set of crop growth records weighted for degree of consistency and grid +.>Crop growth record set with disqualified weighted consistency degree is removed>
According to the fuzzy rough set attribute reduction model, crop growth records with the authority consistency degree failing in the original data subset are removed according to the fuzzy rough set attribute reduction algorithm, so that calculation of interference data is effectively avoided, the calculation cost is reduced, and convenience of the fuzzy rough set attribute reduction algorithm in the actual application process is ensured.
In one embodiment, S4 may include:
calculating the fuzzy similarity of each crop growth record in the original data subset under the crop yield index data through the fuzzy rough set attribute reduction model to obtain a fuzzy set;
Sorting crop growth records in the fuzzy set according to the size of the fuzzy similarity to obtain a weighted fuzzy set;
calculating the weighted consistency degree of each crop growth record in the original data subset with the crop yield grade under the crop yield index data according to the weighted fuzzy set;
removing crop growth records with the weight consistency degree failing in the original data subset according to the weight consistency degree to obtain a crop growth record subset with the weight consistency degree failing in the original data subset;
and calculating the weighted consistency degree and the fuzzy positive area of the crop growth record subset under the crop yield index data according to the fuzzy similarity of each crop growth record in the original data subset under the crop yield index data.
It should be noted that, the network side device may calculate the original data subset (crop sample subset) U through the first calculation of (2) and (3) k Record x of each crop growth i Under the crop yield index data, C= { a 1 ,a 2 ,...,a m Fuzzy similarity to obtain fuzzy set
Wherein x is i 、x j Representing the original data subset U k Crop growth record, a (x) j ) Represents the normalized crop yield index value,representing x under crop yield index a i And x j Fuzzy relation between->Represents x under the index C of the total crop yield i And x j Fuzzy similarity between them, θ represents the calculation parameter.
After obtaining the fuzzy set, the network side equipment can record x according to the crop growth i In fuzzy setIn (2) fuzzy similarity, record x for crop growth i In fuzzy set +.>The similarity of the data is ranked from large to small, and the network side equipment can calculate the original data subset U through the method (4) k Record x of each crop growth i Degree of weighted agreement with crop yield grade D under all crop yield index C>
Wherein,record x representing crop growth i Degree of consistency with weight w, +.>Its calculation is mainly implemented by using fuzzy set +.>The fuzzy similarity of each member is processed by descending order sequencing calculation to form the ordinal number {1,2,3, …, n }, when +.>Is w, then w j =1/k. Setting a weighted consistency degree threshold value eta, and enabling the original data subset U to be subjected to the size of the weighted consistency degree k Crop growth record set divided into weighted consistency degree inequality>And a set of crop growth records weighted for degree of consistency and grid +.>Crop growth record set with disqualified weighted consistency degree is removed>Crop growth record set for obtaining weighted consistency degree and grid >
When (when)When the value of (a) is smaller than a preset threshold value eta (for example, 0.9), judging the crop growth record x i For the crop growth records with the weight consistency degree failing to check, the crop growth records with the weight consistency degree failing to check are removed from the original data subset according to the weight consistency degree, and crop growth records U 'with the weight consistency degree failing to check are obtained' k . The removed crop growth records do not participate in the subsequent attribute reduction step, thereby effectively reducing some inconsistent samplesAnd the calculation efficiency and the accuracy are improved.
Further, in obtaining the weighted consistency degree and the qualification crop growth record subset U' k Thereafter, the network side device may be based on fuzzy setsDividing the weighted consistency degree and the crop growth record subset U 'of the grids according to the upper and lower approximation type (5)' k Fuzzy positive region +.>
In the method, in the process of the invention,inside->Representation->The similarity of the two layers is less than or equal to x i ] D I.e. Inside->Representing two fuzzy sets +.>Wherein the empty set in the fuzzy set is +.>0/x n Is the zade notation in fuzzy set, which represents x n Is 0, [ x ] i ] D Represents x i A set of objects at crop yield level (belonging to the set if the decision attribute of the object is the same as x, or not belonging to the set otherwise), -a set of objects at crop yield level >Representing satisfaction->All x of this condition i Is U 'of' k Crop growth record subset, x, representing weighted degree of consistency and grid i Represents crop growth record, D represents crop yield grade,/->Crop growth record subset U 'representing weighted degree of consistency and grid' k The blurred positive area under all crop yield index C.
In step S5, the network side device calculates a weighted consistency degree and a subset U 'of crop growth records according to the decision table' k Importance Sig of crop yield index α And (a, red, D, U '), a epsilon C, performing attribute reduction on the crop yield index according to the importance of the crop yield index, and selecting the optimal crop yield index a epsilon C until a stopping condition is met to obtain a reduced attribute subset red'.
In one embodiment, S5 may include:
initializing an original reduced attribute set;
crop growth record subset U 'based on weighted degree of consistency and grid' k Calculating a weighted consistency degree and a qualified crop growth record subset U 'of the fuzzy positive area under all crop yield indexes C' k Obtaining an updated crop growth record subset from the fuzzy positive region under the original reduced attribute set;
crop growth record subset U 'based on weighted degree of consistency and grid' k Calculating a weighted consistency degree and a qualified crop growth record subset U 'of the fuzzy positive area under all crop yield indexes C' k Importance Sig of crop yield index α (a,red,D,U′),a∈C；
And according to the importance of the crop yield index, performing attribute reduction on the crop yield index, and selecting the optimal crop yield index a epsilon C until a stopping condition is reached to obtain a reduced attribute subset red'.
It should be noted that, the network side device may initialize the original reduction attribute set red= { } (red is the original reduction set representing the crop yield index, the reduction set of the crop yield index is empty, and then the important index is continuously selected, so as to continuously increase. Then according to the original reduced attribute set red and the weighted consistency degree and the crop growth record subset U 'of the grid' k Fuzzy positive region at all crop yield index CObtaining a weighted consistency degree and a qualified crop growth record subset U 'through the formula (7)' k Fuzzy positive region +.>
Satisfy->
Obtaining updated crop growth record subsets From U' k ×U′ k Reduced to U * ×U′ k 。
The principle of the formula (7) is the same as that of the formula (5), wherein, Inside->Representation->The membership degree in the matrix is less than or equal to x i ] D I.e. +.> Inside->Representing two fuzzy setsWherein the empty set in the fuzzy set is +.>0/x n Is the zade notation in fuzzy set, which represents x n Is 0, [ x ] i ] D Represents x i A set of objects at the crop level,the representation satisfies/>All x of this condition i Set of->Representing satisfaction->All x of this condition i Is U 'of' k Crop growth record set, x, representing weighted degree of consistency and grid i Represents crop growth record, D represents crop yield grade,/->Crop growth record subset U 'representing weighted degree of consistency and grid' k The blurred positive region under the original reduced attribute set red.
Further, the network side equipment can select each crop yield index a epsilon B, calculate and update the crop growth record subset U through the formula (9) * Importance Sig of crop yield index α (a,red,D,U * )。
In Sig α (a,red,D,U * ) Representing an updated crop growth record subset U * Importance of crop yield index, gamma α (red∪{a},D,U * ) Representing a subset U of updated crop growth records based on current * Deterministic metric function, γ, for a given set of attributes red { a }, U } α (red,D,U * ) Representing a subset U of updated crop growth records based on current * Deterministic metric function, x, for a given set of attributes red i A record of the growth of the crop is shown,represents x i In red U-shapeda set of fuzzy membership degrees, [ x ] i ] D Represents x i Collections of objects (crop growth records) at crop level, +.>Represents x i A set of fuzzy membership under red, < +.>Represents x i Fuzzy similarity (fuzzy membership) in the set of fuzzy membership under red U.a, [ x ] i ] D (x j ) Represents 0 or 1, x i And x j The crop grade of (2) is 1, otherwise 0,/or->Represents x i Fuzzy similarity (fuzzy membership) in the set of fuzzy membership under red.
The network side device may continuously select an optimal crop yield index, for example, the crop yield index with the greatest importance of the crop yield index is subjected to attribute reduction by formulas (10) - (12) to update the crop yield index set and the original reduced attribute set.
Sig α (a*,red,D,U * )＝max{Sig α (a k ,red,D,U * ),a k ∈B}(10),
B＝B-{a*}(11),
red＝red∪a*(12),
Where a represents the crop yield index (attribute) with the optimal importance (sig maximum value) of the crop yield index, red represents the original about Jian Jige of the crop yield index, the reduced set of the crop yield index is empty just before the start, and then the important index is continuously selected, so that the number is continuously increased. B is initialized to the overall crop yield index C, and each time one index a e C is selected to enter red, B will delete a, and thus red u b=c. D represents crop yield grade, U * Representing an updated subset of crop growth records, B 'representing an updated set of crop yield indicators, red' representing an updated original reductionA sexual collection.
In the process of attribute reduction, it is necessary to determine whether a stop condition is reached, and a final reduced attribute subset is obtained. Different stop conditions are used for different situations.
When (when)When using stop condition 1:
Sig α (a,red,D,U * ) Delta, delta=0.005, where Sig α (a,red,D,U * ) The importance of the crop yield index of the crop growth record subset is updated after the original reduced attribute set is updated, delta represents a first preset parameter and can be adjusted according to actual conditions.
When (when)When using stop condition 2: />
Wherein the method comprises the steps ofRepresenting a subset of crop growth records based on the current crop>Deterministic metric function given the set of attributes red { a }, ∈>A record of crop growth which indicates entry into the fuzzy positive region under the entire crop yield index set C, i.e./I>Represents x i Set of fuzzy membership under red [. Times.x ] i ] D Represents x i A collection of objects (crop growth records) at the crop level,phi represents a second preset parameter phi= [0,1 ]]The current algorithm may take 0.95 according to the actual adjustment.
If the fuzzy rough set attribute reduction algorithm reaches the stop condition, stopping selecting the optimal product index, ending the attribute reduction step, otherwise, continuing selecting the optimal index until the stop condition is reached or Until that point.
In step S6, the network side device performs redundancy check on the about Jian Shuxing subset to obtain a non-redundant reduced attribute subset.
In one embodiment, S6 may include:
screening out redundancy attributes from the approximately Jian Shuxing subset according to redundancy check conditions;
and eliminating the redundant attribute from the reduced attribute subset to obtain a non-redundant reduced attribute subset.
The redundancy check condition is specifically as follows:
fuzzy positive regions under crop yield index data for a subset of crop growth records when the weighted degree of consistency is acceptableAnd when one of the about Jian Shuxing subsets is removed about Jian Shuxing a epsilon red', gamma is satisfied α (red'\{a},D,U′)＝γ α (C, D, U') and judging the rejected reduced attribute as a redundant attribute.
That is, the network side device can determine the time by judgingBy if a epsilon red' satisfies the condition gamma α (red'\{a},D,U′)＝γ α (C, D, U '), then a may be considered a redundant attribute, reducing the redundant attribute of the reduced attribute subset by red "=red' \ { a }, resulting in a reduced non-redundant reduced attribute subset red".
In step S7, the network side device sequentially inputs the remaining crop growth record data subsets as incremental data subsets into the fuzzy rough set attribute reduction model, updates the original data subsets, repeats steps S4 to S6 with the updated original data subsets to obtain a plurality of non-redundant reduced attribute subsets, and inputs the non-redundant reduced attribute subsets into the classifier model for training to obtain a crop yield prediction model.
It should be noted that, the classifier model may be an SVM support vector machine model, or may be replaced by a model such as a decision tree, a random forest, or a neural network.
According to the crop yield prediction method based on the increment fuzzy rough set attribute reduction, provided by the embodiment of the application, the crop growth record data is divided into a plurality of crop growth record data subsets with equal sizes by constructing the decision table, so that the data volume for attribute reduction each time can be reduced, the technical difficulty is reduced, and the calculation time is shortened; when the attribute is reduced, one crop growth record data subset is used as an original data subset to be input into a fuzzy rough set attribute reduction model for first attribute reduction, the rest crop growth record data subset is used as an incremental data subset to be sequentially input into the fuzzy rough set attribute reduction model, next attribute reduction is performed after the original data subset is updated until the incremental data subset is empty, technical difficulty in the calculation process is reduced, calculation efficiency is improved, more accurate attribute reduction results can be obtained, and accuracy of a crop yield prediction model is improved; redundancy check is carried out on the reduced attribute subset after attribute reduction is carried out each time, and under the condition that the attribute reduction calculation is carried out by increasing the incremental data subsets, the simplified non-redundant reduced attribute subset can be ensured to be obtained, and further long-term effectiveness and reliability of the algorithm under a dynamic environment are ensured.
According to the crop yield prediction method based on the increment fuzzy rough set attribute reduction, firstly, the fuzzy rough set is used for describing crop growth record data, the actual crop growth information yield prediction process is met, the information quantity of the original data is reserved, and certain information loss is avoided. Moreover, the crop yield prediction method based on the increment fuzzy rough set attribute reduction provided by the embodiment of the application is based on fuzzy similarity, utilizes a fuzzy rough set model to analyze inconsistent data in data, effectively avoids the calculation of interference data, reduces the calculation cost, effectively integrates consistent data, further reduces the calculation cost and ensures the algorithm to be fast and convenient in practical application.
Furthermore, the crop yield prediction method based on the increment fuzzy rough set attribute reduction provided by the embodiment of the application constructs an increment fuzzy rough set acceleration model based on an increment learning theory, the model meets the characteristic that data can be continuously calculated even if the data is continuously added, the newly added data can be calculated while the reduction result of historical data is maintained, the static calculation cost is effectively avoided, and the increment learning can calculate a large-scale data set, so that the problem of operation memory is avoided.
The crop yield prediction device based on the reduction of the increment fuzzy rough set attribute provided by the embodiment of the application is described below, and the crop yield prediction device based on the reduction of the increment fuzzy rough set attribute described below and the crop yield prediction method based on the reduction of the increment fuzzy rough set attribute described above can be correspondingly referred to each other.
Fig. 2 is a schematic structural diagram of a crop yield prediction device based on delta fuzzy rough set attribute reduction according to an embodiment of the present application. Referring to fig. 2, an embodiment of the present application provides a crop yield prediction apparatus based on delta fuzzy rough set attribute reduction, which may include:
the decision table construction module 210 is configured to execute step S1: acquiring crop growth record data, crop yield index data and crop yield grade data, constructing a decision table, and setting a fuzzy equivalence relation and a fuzzy similarity threshold;
the crop growth record data dividing module 220 is configured to perform step S2: dividing the crop growth record data into a plurality of crop growth record data subsets with equal sizes according to the decision table;
the first input module 230 is configured to perform step S3: inputting one of a plurality of crop growth record data subsets as an original data subset into a fuzzy rough set attribute reduction model constructed based on a decision table, a fuzzy equivalence relation and a fuzzy similarity threshold;
The crop growth record rejection module 240 is configured to perform step S4: calculating fuzzy similarity of each crop growth record in the original data subset under crop yield index data through the fuzzy rough set attribute reduction model, calculating weighted consistency degree of each crop growth record in the original data subset according to the fuzzy similarity, and eliminating crop growth records with unqualified consistency degree in the original data subset to obtain a crop growth record subset with weighted consistency degree and qualified consistency degree;
attribute No. Jian Mokuai, for performing step S5: according to the decision table, calculating the weighted consistency degree and the importance of the crop yield indexes of the crop growth record subset of the grid, and carrying out attribute reduction on the crop yield indexes according to the importance of the crop yield indexes to obtain a reduced attribute subset;
the redundancy check module 260 is configured to perform step S6: performing redundancy check on the reduced Jian Shuxing subset to obtain a non-redundant reduced attribute subset;
a loop module 270, configured to perform step S7: and sequentially inputting the rest crop growth record data subsets serving as incremental data subsets into the fuzzy rough set attribute reduction model, updating the original data subsets, repeating the steps S4 to S6 by using the updated original data subsets to obtain a plurality of non-redundant reduction attribute subsets, and inputting the non-redundant reduction attribute subsets into the classifier model to train to obtain the crop yield prediction model.
In one embodiment, the decision table construction module 210 may include:
the crop yield grade data obtaining module is used for: acquiring crop growth record data, and clustering the crop growth record data according to crop yield to obtain crop yield grade data;
the crop yield index data obtaining module is used for: normalizing the crop yield index value of the crop growth record data to obtain crop yield index data;
a construction module for: and constructing a decision table DT= (U, C U D) according to the crop growth record data, the crop yield index data and the crop yield grade data, wherein DT represents the decision table, U represents the crop growth record set, C represents the crop yield index set, and D represents the crop yield grade set.
In one embodiment, the crop growth record culling module 240 may include:
a fuzzy set obtaining module for: calculating the fuzzy similarity of each crop growth record in the original data subset under the crop yield index data through the fuzzy rough set attribute reduction model to obtain a fuzzy set;
the weighted fuzzy set obtaining module is used for: sorting crop growth records in the fuzzy set according to the size of the fuzzy similarity to obtain a weighted fuzzy set;
The weighted consistency degree calculation module is used for: calculating the weighted consistency degree of each crop growth record in the original data subset with the crop yield grade under the crop yield index data according to the weighted fuzzy set;
the rejecting module is used for: and removing crop growth records with the weight consistency degree failing in the original data subset according to the weight consistency degree to obtain a crop growth record subset with the weight consistency degree failing.
In one embodiment, the crop growth record culling module 240 may further include:
a fuzzy positive region first calculation module for: and calculating the weighted consistency degree and the fuzzy positive area of the crop growth record subset under the crop yield index data according to the fuzzy similarity of each crop growth record in the original data subset under the crop yield index data.
In one embodiment, the attribute reduction module 250 may include:
an initialization module for: initializing an original reduced attribute set;
a fuzzy positive region second calculation module for: calculating the fuzzy positive areas of the weighted consistency degree and the grid crop growth record subsets under the original reduced attribute sets according to the fuzzy positive areas of the weighted consistency degree and the grid crop growth record subsets under the crop yield index data, and obtaining updated crop growth record subsets;
The index importance calculating module is used for: calculating the importance of crop yield indexes of the updated crop growth record subset according to the weighted consistency degree and the fuzzy positive area of the crop growth record subset under the original reduced attribute set;
the attribute reduction calculation module is used for: and according to the importance of the crop yield index, carrying out attribute reduction on the crop yield index until a stopping condition is reached, so as to obtain a reduced attribute subset.
In one embodiment, the redundancy check module 260 may include:
redundancy attribute screening module for: screening out redundancy attributes from the approximately Jian Shuxing subset according to redundancy check conditions;
the redundant attribute rejecting module is used for: and eliminating the redundant attribute from the reduced attribute subset to obtain a non-redundant reduced attribute subset.
The redundancy check condition is specifically as follows:
and when the fuzzy positive area of the crop growth record subset with the weighted consistency degree and the grid is larger than zero under the crop yield index data, and the crop yield index importance after one of the reduced attributes is removed from the about Jian Shuxing subset is equal to the crop yield index importance of the crop yield index data, judging that the removed reduced attribute is a redundant attribute.
Fig. 3 illustrates a physical schematic diagram of an electronic device, as shown in fig. 3, where the electronic device may include: processor 810, communication interface (Communication Interface) 820, memory 830, and communication bus 840, wherein processor 810, communication interface 820, memory 830 accomplish communication with each other through communication bus 840. Processor 810 may call a computer program in memory 830 to perform the steps of a crop yield prediction method based on delta fuzzy rough set attribute reduction, including, for example:
step S1: acquiring crop growth record data, crop yield index data and crop yield grade data, constructing a decision table, and setting a fuzzy equivalence relation and a fuzzy similarity threshold;
step S2: dividing the crop growth record data into a plurality of crop growth record data subsets with equal sizes according to the decision table;
step S3: inputting one of a plurality of crop growth record data subsets as an original data subset into a fuzzy rough set attribute reduction model constructed based on a decision table, a fuzzy equivalence relation and a fuzzy similarity threshold;
step S4: calculating fuzzy similarity of each crop growth record in the original data subset under crop yield index data through the fuzzy rough set attribute reduction model, calculating weighted consistency degree of each crop growth record in the original data subset according to the fuzzy similarity, and eliminating crop growth records with unqualified consistency degree in the original data subset to obtain a crop growth record subset with weighted consistency degree and qualified consistency degree;
Step S5: according to the decision table, calculating the weighted consistency degree and the importance of the crop yield indexes of the crop growth record subset of the grid, and carrying out attribute reduction on the crop yield indexes according to the importance of the crop yield indexes to obtain a reduced attribute subset;
step S6: performing redundancy check on the reduced Jian Shuxing subset to obtain a non-redundant reduced attribute subset;
step S7: and sequentially inputting the rest crop growth record data subsets serving as incremental data subsets into the fuzzy rough set attribute reduction model, updating the original data subsets, repeating the steps S4 to S6 by using the updated original data subsets to obtain a plurality of non-redundant reduction attribute subsets, and inputting the non-redundant reduction attribute subsets into the classifier model to train to obtain the crop yield prediction model.
Further, the logic instructions in the memory 830 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on this understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a usb disk, a removable hard disk, a Read-only memory (ROM), a random access memory (RAM, randomAccessMemory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In another aspect, embodiments of the present application further provide a computer program product, where the computer program product includes a computer program, where the computer program is capable of being stored on a non-transitory computer readable storage medium, and when the computer program is executed by a processor, the computer is capable of executing the steps of the crop yield prediction method based on the reduction of the incremental fuzzy rough set attribute provided in the above embodiments, for example, including:
step S1: acquiring crop growth record data, crop yield index data and crop yield grade data, constructing a decision table, and setting a fuzzy equivalence relation and a fuzzy similarity threshold;
step S2: dividing the crop growth record data into a plurality of crop growth record data subsets with equal sizes according to the decision table;
step S3: inputting one of a plurality of crop growth record data subsets as an original data subset into a fuzzy rough set attribute reduction model constructed based on a decision table, a fuzzy equivalence relation and a fuzzy similarity threshold;
step S4: calculating fuzzy similarity of each crop growth record in the original data subset under crop yield index data through the fuzzy rough set attribute reduction model, calculating weighted consistency degree of each crop growth record in the original data subset according to the fuzzy similarity, and eliminating crop growth records with unqualified consistency degree in the original data subset to obtain a crop growth record subset with weighted consistency degree and qualified consistency degree;
Step S5: according to the decision table, calculating the weighted consistency degree and the importance of the crop yield indexes of the crop growth record subset of the grid, and carrying out attribute reduction on the crop yield indexes according to the importance of the crop yield indexes to obtain a reduced attribute subset;
step S6: performing redundancy check on the reduced Jian Shuxing subset to obtain a non-redundant reduced attribute subset;
step S7: and sequentially inputting the rest crop growth record data subsets serving as incremental data subsets into the fuzzy rough set attribute reduction model, updating the original data subsets, repeating the steps S4 to S6 by using the updated original data subsets to obtain a plurality of non-redundant reduction attribute subsets, and inputting the non-redundant reduction attribute subsets into the classifier model to train to obtain the crop yield prediction model.
In another aspect, embodiments of the present application further provide a processor-readable storage medium storing a computer program for causing a processor to execute the steps of the method provided in the above embodiments, for example, including:
step S1: acquiring crop growth record data, crop yield index data and crop yield grade data, constructing a decision table, and setting a fuzzy equivalence relation and a fuzzy similarity threshold;
Step S2: dividing the crop growth record data into a plurality of crop growth record data subsets with equal sizes according to the decision table;
step S3: inputting one of a plurality of crop growth record data subsets as an original data subset into a fuzzy rough set attribute reduction model constructed based on a decision table, a fuzzy equivalence relation and a fuzzy similarity threshold;
step S4: calculating fuzzy similarity of each crop growth record in the original data subset under crop yield index data through the fuzzy rough set attribute reduction model, calculating weighted consistency degree of each crop growth record in the original data subset according to the fuzzy similarity, and eliminating crop growth records with unqualified consistency degree in the original data subset to obtain a crop growth record subset with weighted consistency degree and qualified consistency degree;
step S5: according to the decision table, calculating the weighted consistency degree and the importance of the crop yield indexes of the crop growth record subset of the grid, and carrying out attribute reduction on the crop yield indexes according to the importance of the crop yield indexes to obtain a reduced attribute subset;
step S6: performing redundancy check on the reduced Jian Shuxing subset to obtain a non-redundant reduced attribute subset;
Step S7: and sequentially inputting the rest crop growth record data subsets serving as incremental data subsets into the fuzzy rough set attribute reduction model, updating the original data subsets, repeating the steps S4 to S6 by using the updated original data subsets to obtain a plurality of non-redundant reduction attribute subsets, and inputting the non-redundant reduction attribute subsets into the classifier model to train to obtain the crop yield prediction model.
The processor-readable storage medium may be any available medium or data storage device that can be accessed by a processor, including, but not limited to, magnetic storage (e.g., floppy disks, hard disks, magnetic tape, magneto-optical disks (MOs), etc.), optical storage (e.g., CD, DVD, BD, HVD, etc.), semiconductor storage (e.g., ROM, EPROM, EEPROM, nonvolatile storage (NAND FLASH), solid State Disk (SSD)), and the like.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present application, and are not limiting; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims (4)

1. A crop yield prediction method based on delta fuzzy rough set attribute reduction, comprising:
step S1: acquiring crop growth record data, crop yield index data and crop yield grade data, constructing a decision table, and setting a fuzzy equivalence relation and a fuzzy similarity threshold;
the step S1 includes:
acquiring crop growth record data, and clustering the crop growth record data according to crop yield to obtain crop yield grade data;
normalizing the crop yield index value of the crop growth record data to obtain crop yield index data;
constructing a decision table according to the crop growth record data, the crop yield index data and the crop yield grade dataWherein->Representing a decision table->Representing a crop growth record set->Indicating a set of crop yield indicators->Representing a set of crop yield levels;
step S2: dividing the crop growth record data into a plurality of crop growth record data subsets with equal sizes according to the decision table;
step S3: inputting one of a plurality of crop growth record data subsets as an original data subset into a fuzzy rough set attribute reduction model constructed based on a decision table, a fuzzy equivalence relation and a fuzzy similarity threshold;
Step S4: calculating fuzzy similarity of each crop growth record in the original data subset under crop yield index data through the fuzzy rough set attribute reduction model, calculating weighted consistency degree of each crop growth record in the original data subset according to the fuzzy similarity, and eliminating crop growth records with unqualified consistency degree in the original data subset to obtain a crop growth record subset with weighted consistency degree and qualified consistency degree;
the step S4 includes:
calculating the fuzzy similarity of each crop growth record in the original data subset under the crop yield index data through the fuzzy rough set attribute reduction model to obtain a fuzzy set;
sorting crop growth records in the fuzzy set according to the size of the fuzzy similarity to obtain a weighted fuzzy set;
calculating the weighted consistency degree of each crop growth record in the original data subset with the crop yield grade under the crop yield index data according to the weighted fuzzy set;
removing crop growth records with weight consistency degree failing to be qualified from the original data subset according to the weight consistency degree to obtain a crop growth record subset with weight consistency degree failing to be qualified;
The step S4 further includes:
calculating the weighted consistency degree and the fuzzy positive area of the crop growth record subset under the crop yield index data according to the fuzzy similarity of each crop growth record in the original data subset under the crop yield index data; calculating the fuzzy similarity of each crop growth record in the original data subset under the crop yield index data through the fuzzy rough set attribute reduction model, and obtaining a fuzzy set comprises the following steps:
computing the raw data subset according to formulas (2) and (3)Record of growth of each crop->Under the crop yield index data->Is to obtain fuzzy set +.>；
（2）；
（3）；
In the method, in the process of the invention,、/>representing the original data subset +.>Crop growth records,>indicating normalized crop yield index value, +.>Indicating the crop yield index->Down->And->Fuzzy relation between->Indicating the index of yield in all crops->Down->And->Fuzzy similarity between->Representing the calculated parameters;
the calculating the weighted consistency degree of each crop growth record in the original data subset with the crop yield grade under the crop yield index data according to the weighted fuzzy set comprises:
Computing the raw data subset according to equation (4)Record x of each crop growth i Index of yield in all crops->Lower and crop yield grade->Degree of weighted consistency->；
（4）；
Representing crop growth record +.>Weighted +.>Degree of consistency of->The method comprises the steps of carrying out a first treatment on the surface of the Its calculation is mainly implemented by using fuzzy set +.>The fuzzy similarity of each member is subjected to descending order sequencing calculation to form ordinal numbers {1,2,3, …,nwhen->Ordinal number of->Then->；
Step S5: according to the decision table, calculating the weighted consistency degree and the importance of the crop yield indexes of the crop growth record subset of the grid, and carrying out attribute reduction on the crop yield indexes according to the importance of the crop yield indexes to obtain a reduced attribute subset;
the step S5 includes:
initializing an original reduced attribute set;
fuzzy positive regions of the subset of crop growth records under crop yield index data according to the weighted degree of consistency and gridCalculating the fuzzy positive area of the crop growth record subset with the weighted consistency degree and the grid under the original reduced attribute set>Obtaining an updated crop growth record subset;
（7）；
satisfy-> （8）；
In the method, in the process of the invention,inside->Representation->The membership degree of the interior is less than or equal to +.>I.e. ；/>Inside->Representing min { { about }, in two fuzzy sets>The number of empty sets in the fuzzy set is +.0 +.>，/>Is the zade notation in fuzzy set, which indicates +.>Is 0, +.>Representation->Object set at crop level, +.>Representing satisfaction->All->Set of->The representation satisfiesAll->Set of->Crop growth record set representing weighted degree of consistency and grid,/->Indicating crop growth record,/->Indicating crop yield grade,/->Crop growth record subset +.>In the original reduction attribute set +.>A lower blurred positive region;
according to the weighted consistency degree and the gridFuzzy positive regions of crop growth record subsets under original reduced attribute setsCalculating the importance of crop yield index of the updated crop growth record subset>；
（9）；
In the method, in the process of the invention,representing an update crop growth record subset +.>The importance of the crop yield index,representing a subset of updated crop growth records based on current +.>In a given attribute set->The following deterministic metric function->Representing a subset of updated crop growth records based on current +.>In a given attribute set->The following deterministic metric function->Indicating crop growth record,/- >Representation->At->A set of fuzzy membership degrees below, +.>Representation->Collections of objects (crop growth records) at crop level, +.>Representation->At the position ofA set of fuzzy membership degrees below, +.>Representation->At->Fuzzy similarity (fuzzy membership) in the set of fuzzy membership under +.>Represents 0 or 1, < >>And->The crop grade of (2) is 1, otherwise 0,/or->Representation->At->Fuzzy similarity in the set of fuzzy membership degrees;
according to the importance of the crop yield index, carrying out attribute reduction on the crop yield index until a stopping condition is reached, so as to obtain a reduced attribute subset;
step S6: performing redundancy check on the reduced Jian Shuxing subset to obtain a non-redundant reduced attribute subset;
the step S6 includes:
screening out redundancy attributes from the approximately Jian Shuxing subset according to redundancy check conditions;
removing the redundant attribute from the reduced attribute subset to obtain a non-redundant reduced attribute subset;
the redundancy check condition is specifically as follows:
when the fuzzy positive area of the crop growth record subset with the weighted consistency degree and the grid is larger than zero under the crop yield index data, and the crop yield index importance after one of the reduced attributes is removed from the about Jian Shuxing subset is equal to the crop yield index importance of the crop yield index data, judging that the removed reduced attribute is a redundant attribute;
Step S7: and sequentially inputting the rest crop growth record data subsets serving as incremental data subsets into the fuzzy rough set attribute reduction model, updating the original data subsets, repeating the steps S4 to S6 by using the updated original data subsets to obtain a plurality of non-redundant reduction attribute subsets, and inputting the non-redundant reduction attribute subsets into the classifier model to train to obtain the crop yield prediction model.
2. A crop yield prediction device based on delta fuzzy rough set attribute reduction, comprising:
the decision table construction module is used for executing step S1: acquiring crop growth record data, crop yield index data and crop yield grade data, constructing a decision table, and setting a fuzzy equivalence relation and a fuzzy similarity threshold;
the step S1 includes:
acquiring crop growth record data, and clustering the crop growth record data according to crop yield to obtain crop yield grade data;
normalizing the crop yield index value of the crop growth record data to obtain crop yield index data;
constructing a decision table according to the crop growth record data, the crop yield index data and the crop yield grade data Wherein->Representing a decision table->Representing a crop growth record set->Indicating a set of crop yield indicators->Representing a set of crop yield levels;
the crop growth record data dividing module is used for executing the step S2: dividing the crop growth record data into a plurality of crop growth record data subsets with equal sizes according to the decision table;
the first input module is used for executing step S3: inputting one of a plurality of crop growth record data subsets as an original data subset into a fuzzy rough set attribute reduction model constructed based on a decision table, a fuzzy equivalence relation and a fuzzy similarity threshold;
the crop growth record rejecting module is used for executing the step S4: calculating fuzzy similarity of each crop growth record in the original data subset under crop yield index data through the fuzzy rough set attribute reduction model, calculating weighted consistency degree of each crop growth record in the original data subset according to the fuzzy similarity, and eliminating crop growth records with unqualified consistency degree in the original data subset to obtain a crop growth record subset with weighted consistency degree and qualified consistency degree;
the step S4 includes:
Calculating the fuzzy similarity of each crop growth record in the original data subset under the crop yield index data through the fuzzy rough set attribute reduction model to obtain a fuzzy set;
sorting crop growth records in the fuzzy set according to the size of the fuzzy similarity to obtain a weighted fuzzy set;
calculating the weighted consistency degree of each crop growth record in the original data subset with the crop yield grade under the crop yield index data according to the weighted fuzzy set;
removing crop growth records with weight consistency degree failing to be qualified from the original data subset according to the weight consistency degree to obtain a crop growth record subset with weight consistency degree failing to be qualified;
the step S4 further includes:
calculating the weighted consistency degree and the fuzzy positive area of the crop growth record subset under the crop yield index data according to the fuzzy similarity of each crop growth record in the original data subset under the crop yield index data; calculating the fuzzy similarity of each crop growth record in the original data subset under the crop yield index data through the fuzzy rough set attribute reduction model, and obtaining a fuzzy set comprises the following steps:
Computing the raw data subset according to formulas (2) and (3)Record of growth of each crop->Under the crop yield index data->Is to obtain fuzzy set +.>；
（2）；
（3）；
In the method, in the process of the invention,、/>representing the original data subset +.>Crop growth records,>indicating normalized crop yield index value, +.>Indicating the yield of cropsMark->Down->And->Fuzzy relation between->Indicating the index of yield in all crops->Down->And->Fuzzy similarity between->Representing the calculated parameters;
the calculating the weighted consistency degree of each crop growth record in the original data subset with the crop yield grade under the crop yield index data according to the weighted fuzzy set comprises:
computing the raw data subset according to equation (4)Record x of each crop growth i Index of yield in all crops->Lower and crop yield grade->Degree of weighted consistency->；
（4）；
Representing crop growth record +.>Weighted +.>Degree of consistency of->；
The attribute reduction module is configured to execute step S5: according to the decision table, calculating the weighted consistency degree and the importance of the crop yield indexes of the crop growth record subset of the grid, and carrying out attribute reduction on the crop yield indexes according to the importance of the crop yield indexes to obtain a reduced attribute subset;
The step S5 includes:
initializing an original reduced attribute set;
fuzzy positive regions of the subset of crop growth records under crop yield index data according to the weighted degree of consistency and gridCalculating the fuzzy positive area of the crop growth record subset with the weighted consistency degree and the grid under the original reduced attribute set>Obtaining an updated crop growth record subset;
（7）；
satisfy-> （8）；
In the method, in the process of the invention,inside->Representation->The membership degree of the interior is less than or equal to +.>I.e.；/>Inside->Representing min { { about }, in two fuzzy sets>The number of empty sets in the fuzzy set is +.0 +.>，/>Is the zade notation in fuzzy set, which indicates +.>Is 0, +.>Representation->Object set at crop level, +.>Representing satisfaction->All->Set of->The representation satisfiesAll->Set of->Crop growth record set representing weighted degree of consistency and grid,/->Indicating crop growth record,/->Indicating crop yield grade,/->Crop growth record representing degree of weighted consistency and gridSubset->In the original reduction attribute set +.>A lower blurred positive region;
fuzzy positive regions of crop growth record subsets under original reduced attribute sets according to weighted consistency degree and grid Calculating the importance of crop yield index of the updated crop growth record subset>；
（9）；
In the method, in the process of the invention,representing an update crop growth record subset +.>The importance of the crop yield index,representing a subset of updated crop growth records based on current +.>In a given attribute set->The following deterministic metric function->The representation is based on the currentUpdating crop growth record subset->In a given attribute set->The following deterministic metric function->Indicating crop growth record,/->Representation->At->A set of fuzzy membership degrees below, +.>Representation->Collections of objects (crop growth records) at crop level, +.>Representation->At the position ofA set of fuzzy membership degrees below, +.>Representation->At->Fuzzy similarity (fuzzy membership) in the set of fuzzy membership under +.>Represents 0 or 1, < >>And->The crop grade of (2) is 1, otherwise 0,/or->Representation->At->Fuzzy similarity in the set of fuzzy membership degrees;
according to the importance of the crop yield index, carrying out attribute reduction on the crop yield index until a stopping condition is reached, so as to obtain a reduced attribute subset;
the redundancy check module is used for executing step S6: performing redundancy check on the reduced Jian Shuxing subset to obtain a non-redundant reduced attribute subset;
The step S6 includes:
screening out redundancy attributes from the approximately Jian Shuxing subset according to redundancy check conditions;
removing the redundant attribute from the reduced attribute subset to obtain a non-redundant reduced attribute subset;
the redundancy check condition is specifically as follows:
when the fuzzy positive area of the crop growth record subset with the weighted consistency degree and the grid is larger than zero under the crop yield index data, and the crop yield index importance after one of the reduced attributes is removed from the about Jian Shuxing subset is equal to the crop yield index importance of the crop yield index data, judging that the removed reduced attribute is a redundant attribute;
a loop module, configured to execute step S7: and sequentially inputting the rest crop growth record data subsets serving as incremental data subsets into the fuzzy rough set attribute reduction model, updating the original data subsets, repeating the steps S4 to S6 by using the updated original data subsets to obtain a plurality of non-redundant reduction attribute subsets, and inputting the non-redundant reduction attribute subsets into the classifier model to train to obtain the crop yield prediction model.
3. An electronic device comprising a processor and a memory storing a computer program, wherein the processor when executing the computer program implements the crop yield prediction method based on delta fuzzy rough set attribute reduction of claim 1.
4. A computer program product comprising a computer program which, when executed by a processor, implements the crop yield prediction method based on delta fuzzy rough set attribute reduction of claim 1.
CN202211066091.XA 2022-09-01 2022-09-01 Crop yield prediction method based on increment fuzzy rough set attribute reduction Active CN115392582B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211066091.XA CN115392582B (en) 2022-09-01 2022-09-01 Crop yield prediction method based on increment fuzzy rough set attribute reduction

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211066091.XA CN115392582B (en) 2022-09-01 2022-09-01 Crop yield prediction method based on increment fuzzy rough set attribute reduction

Publications (2)

Publication Number Publication Date
CN115392582A CN115392582A (en) 2022-11-25
CN115392582B true CN115392582B (en) 2023-11-14

Family

ID=84124546

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211066091.XA Active CN115392582B (en) 2022-09-01 2022-09-01 Crop yield prediction method based on increment fuzzy rough set attribute reduction

Country Status (1)

Country Link
CN (1) CN115392582B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117111585B (en) * 2023-09-08 2024-02-09 广东工业大学 Numerical control machine tool health state prediction method based on tolerance sub-relation rough set

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103632051A (en) * 2013-11-27 2014-03-12 电子科技大学 Data analyzing method of attribute reduction based on conditional entropy of fuzzy rough set
CN104063230A (en) * 2014-07-09 2014-09-24 中国科学院重庆绿色智能技术研究院 Rough set parallel reduction method, device and system based on MapReduce
CN105892287A (en) * 2016-05-09 2016-08-24 河海大学常州校区 Crop irrigation strategy based on fuzzy judgment and decision making system
CN109828996A (en) * 2018-12-21 2019-05-31 西安交通大学 A kind of Incomplete data set rapid attribute reduction
CN110097934A (en) * 2018-12-29 2019-08-06 蚌埠医学院 A kind of attributive character reduction method of electrocardio Ontological concept

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3057005A1 (en) * 2015-02-12 2016-08-17 Maurice Duault Process for computing a score for a search engine used for accessing a database of real estate properties

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103632051A (en) * 2013-11-27 2014-03-12 电子科技大学 Data analyzing method of attribute reduction based on conditional entropy of fuzzy rough set
CN104063230A (en) * 2014-07-09 2014-09-24 中国科学院重庆绿色智能技术研究院 Rough set parallel reduction method, device and system based on MapReduce
CN105892287A (en) * 2016-05-09 2016-08-24 河海大学常州校区 Crop irrigation strategy based on fuzzy judgment and decision making system
CN109828996A (en) * 2018-12-21 2019-05-31 西安交通大学 A kind of Incomplete data set rapid attribute reduction
CN110097934A (en) * 2018-12-29 2019-08-06 蚌埠医学院 A kind of attributive character reduction method of electrocardio Ontological concept

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于粗糙集的农机化发展水平评价方法;郭婉;吴昊;王安宁;易涛;勾志营;叶新玉;;农机化研究(第02期);全文 *

Also Published As

Publication number Publication date
CN115392582A (en) 2022-11-25

Similar Documents

Publication Publication Date Title
CN109350032B (en) Classification method, classification system, electronic equipment and storage medium
CN111524606A (en) Tumor data statistical method based on random forest algorithm
CN111986811A (en) Disease prediction system based on big data
Kianmehr et al. Fuzzy clustering-based discretization for gene expression classification
CN115392582B (en) Crop yield prediction method based on increment fuzzy rough set attribute reduction
CN108682457A (en) Patient&#39;s long-term prognosis quantitative forecast and interfering system and method
Islam et al. Sgbba: An efficient method for prediction system in machine learning using imbalance dataset
Tiruneh et al. Feature selection for construction organizational competencies impacting performance
CN114003900A (en) Network intrusion detection method, device and system for secondary system of transformer substation
CN114242178A (en) Method for quantitatively predicting biological activity of ER alpha antagonist based on gradient lifting decision tree
CN110109005B (en) Analog circuit fault testing method based on sequential testing
CN116881841A (en) Hybrid model fault diagnosis method based on F1-score multistage decision analysis
CN116090618A (en) Operation situation sensing method and device for power communication network
Xu et al. Hybrid label noise correction algorithm for medical auxiliary diagnosis
CN112382382B (en) Cost-sensitive integrated learning classification method and system
Akash et al. Predicting Thyroid Dysfunction Using Machine Learning Techniques
CN115409050A (en) Bearing fault diagnosis method, device and system
CN114363004A (en) Risk assessment method and device, computer equipment and storage medium
CN112768090A (en) Filtering system and method for chronic disease detection and risk assessment
CN112732690A (en) Stabilizing system and method for chronic disease detection and risk assessment
Nancy et al. Optimal feature selection for classification using rough set-based CGA-NN classifier
CN113221966A (en) Differential privacy decision tree construction method based on F _ Max attribute measurement
Shah et al. Analyzing data complexity using metafeatures for classification algorithm selection
Marwah et al. Lung Cancer Survivability prediction with Recursive Feature Elimination using Random Forest and Ensemble Classifiers
CN116226629B (en) Multi-model feature selection method and system based on feature contribution

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant