CN116150610B - Training method, system, computer and storage medium for suspicious error data processing model - Google Patents
Training method, system, computer and storage medium for suspicious error data processing model Download PDFInfo
- Publication number
- CN116150610B CN116150610B CN202310425084.2A CN202310425084A CN116150610B CN 116150610 B CN116150610 B CN 116150610B CN 202310425084 A CN202310425084 A CN 202310425084A CN 116150610 B CN116150610 B CN 116150610B
- Authority
- CN
- China
- Prior art keywords
- error data
- target
- training
- suspicious error
- suspicious
- 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
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Abstract
The invention provides a training method, a training system, a training computer and a training storage medium for a suspicious error data processing model, wherein the training method comprises the following steps: acquiring original historical suspicious error data, and classifying the original historical suspicious error data according to the observation elements to generate a plurality of corresponding characteristic data tables, wherein each characteristic data table comprises a target value and a plurality of characteristic values; adding labeling values to the target value and the characteristic values in sequence to generate a corresponding target characteristic data table, and storing the target characteristic data table into a target document; training a preset decision tree machine learning algorithm based on a target document, and training a corresponding decision tree model so that the decision tree model processes suspicious error data generated in real time. According to the method, the processing and analyzing process of the suspicious error data by manual operation can be omitted, so that the suspicious error data can be automatically and comprehensively processed, and the processing efficiency of the suspicious error data is greatly improved.
Description
Technical Field
The present invention relates to the field of data processing technologies, and in particular, to a training method, a training system, a training computer, and a training storage medium for a suspicious error data processing model.
Background
Most of the existing meteorological departments use a meteorological data service system for recording and storing meteorological data in real time, and accordingly the processing efficiency of the meteorological data is improved.
In order to avoid deviation of meteorological results, the existing meteorological data service system needs to discover and process the generated suspicious erroneous data in time, however, most of the prior art processes the suspicious erroneous data which can be discovered in a manual mode, so that incomplete suspicious erroneous data processing is easy to occur, and deviation is further caused to the meteorological results.
Disclosure of Invention
Based on the above, the invention aims to provide a training method, a training system, a training computer and a training storage medium for a suspicious error data processing model, so as to solve the problems that most of the suspicious error data which can be found is processed manually in the prior art, so that incomplete suspicious error data processing is easy to occur, and deviation is brought to meteorological results.
An embodiment of the present invention provides a training method for a suspicious error data processing model, where the method includes:
acquiring original historical suspicious error data, and classifying the original historical suspicious error data according to an observation element to generate a plurality of corresponding characteristic data tables, wherein each characteristic data table comprises a target value and a plurality of characteristic values;
Adding labeling values to the target value and the characteristic values in sequence to generate a corresponding target characteristic data table, and storing the target characteristic data table into a target document;
training a preset decision tree machine learning algorithm based on the target document, and training a corresponding decision tree model so that the decision tree model processes suspicious error data generated in real time.
The beneficial effects of the invention are as follows: the method comprises the steps of obtaining original historical suspicious error data, and classifying the original historical suspicious error data according to observation factors to generate a plurality of corresponding characteristic data tables, wherein each characteristic data table comprises a target value and a plurality of characteristic values; further, labeling values are sequentially added to the target value and the characteristic values to generate a corresponding target characteristic data table, and the target characteristic data table is stored in a target document; finally, training a preset decision tree machine learning algorithm based on the target document, and training a corresponding decision tree model to enable the decision tree model to process suspicious error data generated in real time. According to the method, when the weather data service system identifies the suspicious error data, the identified suspicious error data is automatically processed through the trained decision tree model, and the corresponding suspicious error data processing result is generated, so that the manual processing and analyzing process of the suspicious error data is omitted, the suspicious error data can be automatically and comprehensively processed, the processing efficiency of the suspicious error data is greatly improved, and the work efficiency of a weather department is improved.
Preferably, the step of classifying the original historical suspicious error data according to the observation element to generate a plurality of corresponding feature data tables includes:
when the original historical suspicious error data is obtained, detecting observation elements contained in the original historical suspicious error data, and splitting the observation elements to split the original historical suspicious error data into a plurality of corresponding observation element data sets, wherein each observation element corresponds to one observation element data set, and the observation elements comprise air temperature, air pressure, wind direction, wind speed, relative humidity, evaporation capacity, sunlight, visibility and ground temperature;
and identifying a plurality of characteristic values contained in each observation element data set one by one, and predicting a corresponding target value according to the plurality of characteristic values so as to generate the characteristic data table according to the target value and the plurality of characteristic values.
Preferably, the step of identifying a plurality of feature values contained in each observation element data set one by one, and predicting a corresponding target value according to the plurality of feature values includes:
screening out feature information contained in a plurality of feature values in each observation element data set, wherein the feature information comprises quality control result description information, suspicious error type information corresponding to the quality control result description information and observation element information;
Establishing a mapping relation among the quality control result description information, the suspected error type information and the observation element information, wherein the mapping relation has uniqueness, the suspected error type information comprises errors, suspicions and missing tests, and the observation element information comprises observation values;
and predicting a corresponding target processing mode in a preset conclusion database according to the mapping relation, and setting a processing result of the target processing mode as the target value.
Preferably, the step of predicting the corresponding target processing mode in the preset conclusion database according to the mapping relation includes:
when the mapping relation is obtained, a target observation site corresponding to the current observation element data set is found out according to the mapping relation, and a target observation element of the current observation element data set in the target observation site is determined, wherein the target observation site comprises a national station and an regional station;
and predicting the corresponding target processing mode in the preset conclusion database according to the target observation element.
Preferably, the step of sequentially adding a labeling value to the target value and the feature values to generate a corresponding target feature data table includes:
Adding annotation columns at preset positions corresponding to the target value and the characteristic values one by one, and adding corresponding annotation values in the annotation columns one by one according to preset rules, wherein each target value and each annotation value corresponding to the characteristic value have uniqueness;
and respectively establishing the target value and the mapping relation between the characteristic value and the labeling value to generate the target characteristic data table.
Preferably, the step of training a preset decision tree machine learning algorithm based on the target document includes:
initializing the preset decision tree machine learning algorithm, inputting the target document into the initialized decision tree machine learning algorithm, and enabling the initialized decision tree machine learning algorithm to identify suspicious error information and suspicious error data circulation information in the target document;
and performing classification learning on the initialized decision tree machine learning algorithm through the suspicious error information and the suspicious error data circulation information so as to complete training of the preset decision tree machine learning algorithm.
Preferably, the step of performing classification learning on the initialized decision tree machine learning algorithm through the suspicious error information and the suspicious error data flow information to complete training of the preset decision tree machine learning algorithm includes:
When the suspicious error information and the suspicious error data circulation information are respectively acquired, generating a plurality of corresponding training samples according to the suspicious error information, and generating a plurality of corresponding training nodes according to the suspicious error data circulation information, wherein each training sample and each training node have uniqueness;
converting each training sample into a plurality of corresponding feature vectors based on a CART algorithm, and inputting the plurality of feature vectors into a plurality of training nodes correspondingly respectively so as to perform corresponding growth learning on each training node and generate a plurality of corresponding child nodes;
performing continuous growth learning on the plurality of sub-nodes, and judging whether the characteristic attributes corresponding to the plurality of sub-nodes respectively meet preset conditions or not, wherein each sub-node corresponds to one type of characteristic attribute respectively;
if the fact that the characteristic attribute values corresponding to the child nodes respectively meet the preset conditions is detected, the training node is set to be a leaf node, the child nodes with the growth learning completed are set to be non-leaf nodes, and the leaf nodes and the non-leaf nodes are input into the preset decision tree machine learning algorithm at the same time, so that a corresponding target decision tree model is trained.
A second aspect of an embodiment of the present invention provides a training system for a suspicious error data processing model, where the system includes:
the acquisition module is used for acquiring original historical suspicious error data, classifying the original historical suspicious error data according to the observation factors to generate a plurality of corresponding characteristic data tables, wherein each characteristic data table comprises a target value and a plurality of characteristic values;
the labeling module is used for sequentially adding labeling values to the target value and the characteristic values to generate a corresponding target characteristic data table, and storing the target characteristic data table into a target document;
the training module is used for training a preset decision tree machine learning algorithm based on the target document and training a corresponding decision tree model so that the decision tree model can process suspicious error data generated in real time.
In the training system for the suspicious error data processing model, the acquiring module is specifically configured to:
when the original historical suspicious error data is obtained, detecting observation elements contained in the original historical suspicious error data, and splitting the observation elements to split the original historical suspicious error data into a plurality of corresponding observation element data sets, wherein each observation element corresponds to one observation element data set, and the observation elements comprise air temperature, air pressure, wind direction, wind speed, relative humidity, evaporation capacity, sunlight, visibility and ground temperature;
And identifying a plurality of characteristic values contained in each observation element data set one by one, and predicting a corresponding target value according to the plurality of characteristic values so as to generate the characteristic data table according to the target value and the plurality of characteristic values.
In the training system for the suspicious error data processing model, the acquiring module is further specifically configured to:
screening out feature information contained in a plurality of feature values in each observation element data set, wherein the feature information comprises quality control result description information, suspicious error type information corresponding to the quality control result description information and observation element information;
establishing a mapping relation among the quality control result description information, the suspected error type information and the observation element information, wherein the mapping relation has uniqueness, the suspected error type information comprises errors, suspicions and missing tests, and the observation element information comprises observation values;
and predicting a corresponding target processing mode in a preset conclusion database according to the mapping relation, and setting a processing result of the target processing mode as the target value.
In the training system for the suspicious error data processing model, the acquiring module is further specifically configured to:
When the mapping relation is obtained, a target observation site corresponding to the current observation element data set is found out according to the mapping relation, and a target observation element of the current observation element data set in the target observation site is determined, wherein the target observation site comprises a national station and an regional station;
and predicting the corresponding target processing mode in the preset conclusion database according to the target observation element.
In the training system of the suspicious error data processing model, the labeling module is specifically configured to:
adding annotation columns at preset positions corresponding to the target value and the characteristic values one by one, and adding corresponding annotation values in the annotation columns one by one according to preset rules, wherein each target value and each annotation value corresponding to the characteristic value have uniqueness;
and respectively establishing the target value and the mapping relation between the characteristic value and the labeling value to generate the target characteristic data table.
In the training system of the suspicious error data processing model, the training module is specifically configured to:
initializing the preset decision tree machine learning algorithm, inputting the target document into the initialized decision tree machine learning algorithm, and enabling the initialized decision tree machine learning algorithm to identify suspicious error information and suspicious error data circulation information in the target document;
And performing classification learning on the initialized decision tree machine learning algorithm through the suspicious error information and the suspicious error data circulation information so as to complete training of the preset decision tree machine learning algorithm.
In the training system for the suspicious error data processing model, the training system for the suspicious error data processing model further comprises a fusion module, wherein the fusion module is specifically used for:
and carrying out page analysis on the meteorological data service system through a man-machine interaction processing algorithm to sort out execution flow information in the meteorological data service system, and fusing the execution flow information into the decision tree model to generate a corresponding man-machine interaction model.
A third aspect of an embodiment of the present invention proposes a computer comprising a memory, a processor and a computer program stored on said memory and executable on said processor, said processor implementing the above-described method for training a suspicious error data handling model when executing said computer program.
A fourth aspect of the embodiments of the present invention proposes a storage medium having stored thereon a computer program which, when executed by a processor, implements a suspected erroneous data handling model training method as described above.
Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
FIG. 1 is a flowchart of a training method for a suspicious error data handling model according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a decision tree in a training method of a suspicious error data processing model according to an embodiment of the present invention;
FIG. 3 is a flowchart of a predicted target processing method in a training method of a suspicious error data processing model according to an embodiment of the present invention;
FIG. 4 is a block diagram illustrating a training system for a suspicious error data handling model according to one embodiment of the present invention.
The invention will be further described in the following detailed description in conjunction with the above-described figures.
Detailed Description
In order that the invention may be readily understood, a more complete description of the invention will be rendered by reference to the appended drawings. Several embodiments of the invention are presented in the figures. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
It will be understood that when an element is referred to as being "mounted" on another element, it can be directly on the other element or intervening elements may also be present. When an element is referred to as being "connected" to another element, it can be directly connected to the other element or intervening elements may also be present. The terms "vertical," "horizontal," "left," "right," and the like are used herein for illustrative purposes only.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.
In order to avoid deviation of meteorological results, the existing meteorological data service system needs to discover and process the generated suspicious erroneous data in time, however, most of the prior art processes the suspicious erroneous data which can be discovered in a manual mode, so that incomplete suspicious erroneous data processing is easy to occur, and deviation is further caused to the meteorological results.
Referring to fig. 1, a training method for a suspicious and erroneous data processing model according to a first embodiment of the present invention is shown, where when a meteorological data service system identifies suspicious and erroneous data, the training method for a suspicious and erroneous data processing model according to the present embodiment can automatically process the identified suspicious and erroneous data through a trained decision tree model and generate a corresponding suspicious and erroneous data processing result, so that a manual processing and analyzing process for the suspicious and erroneous data is omitted, and therefore, the suspicious and erroneous data can be automatically and comprehensively processed, the processing efficiency of the suspicious and erroneous data is greatly improved, and the work efficiency of a meteorological department is facilitated to be improved.
Specifically, the training method for the suspicious error data processing model provided by the embodiment specifically includes the following steps:
step S10, original historical suspicious error data are obtained, and classification processing is carried out on the original historical suspicious error data according to observation factors so as to generate a plurality of corresponding characteristic data tables, wherein each characteristic data table comprises a target value and a plurality of characteristic values;
specifically, in this embodiment, it should be firstly explained that the training method for the suspicious error data processing model provided in this embodiment is specifically applied to meteorological departments in all regions of the country, and more specifically, specifically applied to meteorological data service systems used by each meteorological department, and is used for automatically implementing processing on suspicious error data identified by the meteorological data service systems, so as to improve processing efficiency on suspicious error data.
The weather data service system can collect and store received weather data in real time, specifically, the weather data comprise correct data and suspicious error data, meanwhile, the weather data service system can also identify the received suspicious error data in real time, and further, in order to avoid the influence of the suspicious error data on the broadcasted weather result, the generated suspicious error data needs to be processed in time.
Therefore, in this step, it should be noted that, in this step, first, the original historical suspicious error data determined in real time by the weather data service system needs to be obtained, and further, in this step, corresponding classification processing is performed on the original historical suspicious error data obtained in real time according to the existing observation elements, that is, the original historical suspicious error data is classified according to the types of the existing observation elements, and a plurality of corresponding feature data tables are generated, where each feature data table corresponds to one observation element. More specifically, each feature data table includes a target value and a number of feature values, wherein the feature values can affect the size of the target value.
Step S20, adding labeling values to the target value and the feature values in sequence to generate a corresponding target feature data table, and storing the target feature data table into a target document;
further, in this embodiment, it should be noted that, in this step, labeling values are further added to the target value and the plurality of feature values in the feature data table in turn, so that a required target feature data table can be further generated, and at the same time, the target feature data table is stored in a preset target document, preferably, in this embodiment, the target document may be a txt document, an excel document, a word document, and the like, which are all within the protection scope of this embodiment.
Step S30, training a preset decision tree machine learning algorithm based on the target document, and training a corresponding decision tree model so that the decision tree model processes suspicious error data generated in real time.
Finally, in this embodiment, it should be noted that, after the required target document is obtained through the above steps, the step further inputs the target document into a preset decision tree machine learning algorithm to train the decision tree machine learning algorithm. Based on the method, a corresponding decision tree model is trained, so that the trained decision tree model can process generated suspicious error data in real time in an actual working process.
When the method is used, the original historical suspicious error data are obtained, and classified according to the observation factors, so that a plurality of corresponding characteristic data tables are generated, wherein each characteristic data table comprises a target value and a plurality of characteristic values; further, labeling values are sequentially added to the target value and the characteristic values to generate a corresponding target characteristic data table, and the target characteristic data table is stored in a target document; finally, training a preset decision tree machine learning algorithm based on the target document, and training a corresponding decision tree model to enable the decision tree model to process suspicious error data generated in real time. According to the method, when the weather data service system identifies the suspicious error data, the identified suspicious error data is automatically processed through the trained decision tree model, and the corresponding suspicious error data processing result is generated, so that the manual processing and analyzing process of the suspicious error data is omitted, the suspicious error data can be automatically and comprehensively processed, the processing efficiency of the suspicious error data is greatly improved, and the work efficiency of a weather department is improved.
It should be noted that the foregoing implementation procedure is only for illustrating the feasibility of the present application, but this does not represent that the suspicious error data processing model training method of the present application has only one implementation procedure, and instead, the suspicious error data processing model training method of the present application can be implemented only by the method, which is included in the feasible embodiments of the present application.
In summary, the training method for the suspicious error data processing model provided by the embodiment of the invention can automatically process the identified suspicious error data through the trained decision tree model when the suspicious error data is identified by the meteorological data service system, and generate a corresponding suspicious error data processing result, thereby omitting the manual processing and analyzing process of the suspicious error data, automatically and comprehensively processing the suspicious error data, greatly improving the processing efficiency of the suspicious error data and being beneficial to improving the working efficiency of meteorological departments.
The second embodiment of the present invention also provides a training method for a suspicious error data processing model, where the training method for a suspicious error data processing model provided in the present embodiment is different from the training method for a suspicious error data processing model provided in the first embodiment in that:
Specifically, in this embodiment, it should be noted that, the step of classifying the original historical suspicious error data according to the observation element to generate a plurality of corresponding feature data tables includes:
when the original historical suspicious error data is obtained, detecting observation elements contained in the original historical suspicious error data, and splitting the observation elements to split the original historical suspicious error data into a plurality of corresponding observation element data sets, wherein the observation elements comprise air temperature, air pressure, wind direction, wind speed, relative humidity, evaporation capacity, sunlight, visibility, ground temperature and the like;
and identifying a plurality of characteristic values contained in each observation element data set one by one, and predicting a corresponding target value according to the plurality of characteristic values so as to generate the characteristic data table according to the target value and the plurality of characteristic values.
Specifically, in this embodiment, it should be noted that, after the original historical suspicious error data is obtained in this embodiment, the embodiment detects, in real time, an observation element included in the current original historical suspicious error data, that is, detects whether one or more observation elements of evaporation amount, air temperature, air pressure and wind speed are included in the current original historical suspicious error data, and further performs splitting processing on the current original historical suspicious error data according to the detected kind of the observation element, so that the current original historical suspicious error data can be split into a plurality of corresponding observation element data sets, where each observation element data set represents one kind of observation element.
Further, in this embodiment, a plurality of feature values included in each current observation element dataset are further identified, a corresponding target value is further predicted according to the current plurality of feature values, and finally, a required feature data table is generated according to the obtained feature values and the target value, specifically, the feature data table prepared by taking the evaporation amount as an example in this embodiment is shown in the following table 1, and in this embodiment, the observation element datasets corresponding to other observation elements may be made into similar feature data tables.
TABLE 1
The first columns include the characteristic values selected through screening, namely quality control result description, suspicious error types, precipitation amount of 1 hour at the current time and evaporation amount of 1 hour before and after the current time, and the last column 1 is a target value, namely a processing mode of evaporating suspicious error data in hours.
Further, in this embodiment, it should be noted that the step of identifying, one by one, a plurality of feature values included in each observation element dataset and predicting a corresponding target value according to the plurality of feature values includes:
screening out feature information contained in a plurality of feature values in each observation element data set, wherein the feature information comprises quality control result description information, suspicious error type information corresponding to the quality control result description information and observation element information;
Establishing a mapping relation among the quality control result description information, the suspected error type information and the observation element information, wherein the mapping relation has uniqueness, the suspected error type information comprises errors, suspicions and missing tests, and the observation element information comprises observation values;
and predicting a corresponding target processing mode in a preset conclusion database according to the mapping relation, and setting a processing result of the target processing mode as the target value.
Specifically, in this step, it should be noted that, in this step, quality control result description information, suspicious error type information corresponding to the quality control result description information, and observation element information included in each observation element data set are screened out one by one, and on this basis, a mapping relationship is established between the three, specifically, taking evaporation amount as an example, the observation element information provided in this embodiment may include whether there is precipitation amount at the current time and whether evaporation amount is less than or equal to 1.0 in 1 hour before and after, and in addition, the observation element may also include other cases, which are all within the protection scope of this embodiment.
As shown in table 1 and fig. 2, further, the corresponding target processing mode is further predicted in the preset conclusion database according to the established mapping relationship, specifically, the embodiment uses the obtained observation element information as a judgment basis, and further accurately predicts the corresponding target processing mode according to the judgment result, and at the same time, the processing result of the obtained target processing mode is set as the target value, so that the required characteristic value and the target value can be simply and conveniently obtained.
It should be noted that, the target processing methods provided in this embodiment mainly include three kinds of processing, namely processing according to 0.0, processing according to interpolation, and processing according to missing measurement. In addition, it should be noted that, the target processing manners provided in this embodiment are all available to those skilled in the art from the prior art.
For easy understanding, for example, the quality control result description information provided in this embodiment specifically includes the following four cases:
(1) "precipitation was found 12 hours past at 08, but no precipitation was found at night, and internal consistency check was not passed"; or "fail 'weather phenomenon format is wrong, when weather phenomenon occurs at 20:00, the night bar' of the next day should be counted; precipitation was found 12 hours past at 08, but no precipitation was found at night, and internal consistency check was not passed "
(2) "precipitation was observed in the past 12 hours at 20 hours, but no precipitation was observed in the daytime, and the internal consistency check was not passed"; or "fail 'weather phenomenon format is wrong, when weather phenomenon occurs at 20:00, the night bar' of the next day should be counted; precipitation was observed in the past 12 hours at 20 hours, but no precipitation was observed in the daytime, and internal consistency check was not passed "
(3) "precipitation was observed 12 hours past at 08, but no precipitation was observed at night, and internal consistency checks were not passed; precipitation occurs in the past 12 hours at 20 hours, but no precipitation occurs in the daytime, the internal consistency check is failed, or the weather phenomenon is wrong, and when the weather phenomenon occurs at 20:00, the night bar' of the next day is counted; precipitation is generated in the past 12 hours at 08, but no precipitation phenomenon exists at night, and the internal consistency check is not passed; precipitation was observed at 20 hours past 12 hours, but no precipitation was observed during the day, and no internal consistency check was passed.
(4) Precipitation occurs in the past 12 hours when '20 is not passed, but no precipitation occurs in the daytime, and the internal consistency check' is not passed; 'no precipitation, but no zero minute precipitation (09:58)' and the like; or "fail 'weather phenomenon format is wrong, when weather phenomenon occurs at 20:00, the night bar' of the next day should be counted; precipitation occurs in the past 12 hours when '20 is not passed, but no precipitation occurs in the daytime, and the internal consistency check' is not passed; 'no precipitation, but no zero minute precipitation (09:58)' and the like.
Further, it should be noted that if the precipitation amount of the above-mentioned (1) (2) (3) (4) is less than or equal to 0.3, the weather phenomenon is treated according to "confirm error", and the corresponding data of precipitation amount of more than or equal to 0.1mm in each minute at 20-8 or 8-20 is modified synchronously to be "0.0"; synchronously modifying the data with the precipitation amount of more than or equal to 0.1mm in each hour of corresponding 20-8 hours or 8-20 hours to be 0.0; the precipitation column for the past 12 hours at 8 hours and 20 hours is modified to 0.0 if precipitation exists.
Furthermore, in this embodiment, the current feature values are deleted according to the missing measurement process, so as to avoid the influence on the prediction result.
In addition, the interpolation processing provided in this embodiment calculates the evaporation value in the current hour in real time, specifically, the evaporation value in the previous and next hour is added and then averaged, so that the required evaporation value can be calculated.
In addition, in this embodiment, it should be noted that the step of predicting the corresponding target processing mode in the preset conclusion database according to the mapping relationship includes:
when the mapping relation is obtained, a target observation site corresponding to the current observation element data set is found out according to the mapping relation, and a target observation element of the current observation element data set in the target observation site is determined, wherein the target observation site comprises a national station and an regional station;
and predicting the corresponding target processing mode in the preset conclusion database according to the target observation element.
Specifically, in this step, as shown in fig. 3, after the mapping relationship between the quality control result description information, the suspected error type information corresponding to the quality control result description information, and the observation element information is obtained in the above manner, this step further searches for a target observation site corresponding to the current observation element data set according to the suspected error information in the mapping relationship, where the target observation site may be a national site or an area site, and further determines a target observation element of the current observation element data set in the current target observation site, where the target observation element may be one of the observation elements such as air temperature, air pressure, wind speed, precipitation, and evaporation amount.
Finally, the step can predict a corresponding target processing mode in the preset conclusion database according to the obtained target observation element, and specifically, the target processing mode can be interpolation processing, missing measurement processing and the like.
It should be noted that the foregoing implementation procedure is only for illustrating the feasibility of the present application, but this does not represent that the suspicious error data processing model training method of the present application has only one implementation procedure, and instead, the suspicious error data processing model training method of the present application can be implemented only by the method, which is included in the feasible embodiments of the present application.
In summary, the training method for the suspicious error data processing model provided by the embodiment of the invention can automatically process the identified suspicious error data through the trained decision tree model when the suspicious error data is identified by the meteorological data service system, and generate a corresponding suspicious error data processing result, thereby omitting the manual processing and analyzing process of the suspicious error data, automatically and comprehensively processing the suspicious error data, greatly improving the processing efficiency of the suspicious error data and being beneficial to improving the working efficiency of meteorological departments.
The third embodiment of the present invention also provides a training method for a suspicious error data processing model, where the training method for a suspicious error data processing model provided in the present embodiment is different from the training method for a suspicious error data processing model provided in the first embodiment in that:
specifically, in this embodiment, it should be noted that the step of sequentially adding the labeling values to the target value and the plurality of feature values to generate the corresponding target feature data table includes:
adding annotation columns at preset positions corresponding to the target value and the characteristic values one by one, and adding corresponding annotation values in the annotation columns one by one according to preset rules, wherein each target value and each annotation value corresponding to the characteristic value have uniqueness;
and respectively establishing the target value and the mapping relation between the characteristic value and the labeling value to generate the target characteristic data table.
Specifically, in this embodiment, in order to accurately obtain the required target feature data table, in this embodiment, corresponding annotation columns are added at preset positions corresponding to the target values and feature values one by one in the table 1, and at the same time, corresponding annotation values are added in the annotation columns one by one according to a preset rule, and a mapping relationship between the target values and feature values and the annotation values added in real time is respectively established, so that the required target feature data table can be generated.
Specifically, the target feature data table provided in this embodiment is shown in table 2
TABLE 2
It should be noted that the foregoing implementation procedure is only for illustrating the feasibility of the present application, but this does not represent that the suspicious error data processing model training method of the present application has only one implementation procedure, and instead, the suspicious error data processing model training method of the present application can be implemented only by the method, which is included in the feasible embodiments of the present application.
In summary, the training method for the suspicious error data processing model provided by the embodiment of the invention can automatically process the identified suspicious error data through the trained decision tree model when the suspicious error data is identified by the meteorological data service system, and generate a corresponding suspicious error data processing result, thereby omitting the manual processing and analyzing process of the suspicious error data, automatically and comprehensively processing the suspicious error data, greatly improving the processing efficiency of the suspicious error data and being beneficial to improving the working efficiency of meteorological departments.
The fourth embodiment of the present invention also provides a training method for a suspicious error data processing model, where the training method for a suspicious error data processing model provided in the present embodiment is different from the training method for a suspicious error data processing model provided in the first embodiment in that:
Specifically, in this embodiment, it should also be noted that the step of training the preset decision tree machine learning algorithm based on the target document includes:
specifically, in this embodiment, it should be noted that, in order to accurately train the preset decision tree machine learning algorithm, in this embodiment, the preset decision tree machine learning algorithm is initialized first, and the target document is input into the initialized decision tree machine learning algorithm, so that the initialized decision tree machine learning algorithm identifies suspicious error information and suspicious error data circulation information in the target document;
furthermore, in this embodiment, the decision tree machine learning algorithm after the initialization process is subjected to classification learning through the suspicious error information and the suspicious error data flow information, so as to complete training of the preset decision tree machine learning algorithm.
It should be noted that the foregoing implementation procedure is only for illustrating the feasibility of the present application, but this does not represent that the suspicious error data processing model training method of the present application has only one implementation procedure, and instead, the suspicious error data processing model training method of the present application can be implemented only by the method, which is included in the feasible embodiments of the present application.
In summary, the training method for the suspicious error data processing model provided by the embodiment of the invention can automatically process the identified suspicious error data through the trained decision tree model when the suspicious error data is identified by the meteorological data service system, and generate a corresponding suspicious error data processing result, thereby omitting the manual processing and analyzing process of the suspicious error data, automatically and comprehensively processing the suspicious error data, greatly improving the processing efficiency of the suspicious error data and being beneficial to improving the working efficiency of meteorological departments.
The fifth embodiment of the present invention also provides a training method for a suspicious error data processing model, where the training method for a suspicious error data processing model provided in the present embodiment is different from the training method for a suspicious error data processing model provided in the first embodiment in that:
in addition, in this embodiment, it should be further noted that, the step of performing classification learning on the initialized decision tree machine learning algorithm by using the suspicious error information and the suspicious error data flow information to complete training of the preset decision tree machine learning algorithm includes:
when the suspicious error information and the suspicious error data circulation information are respectively acquired, generating a plurality of corresponding training samples according to the suspicious error information, and generating a plurality of corresponding training nodes according to the suspicious error data circulation information, wherein each training sample and each training node have uniqueness;
Converting each training sample into a plurality of corresponding feature vectors based on a CART algorithm, and inputting the plurality of feature vectors into a plurality of training nodes correspondingly respectively so as to perform corresponding growth learning on each training node and generate a plurality of corresponding child nodes;
performing continuous growth learning on a plurality of sub-nodes, and judging whether characteristic attributes corresponding to the sub-nodes respectively meet preset conditions or not, namely judging whether the characteristic attributes corresponding to each sub-node meet the condition of stopping growth or not, wherein each sub-node respectively corresponds to one type of characteristic attribute, and specifically, the characteristic attributes provided by the embodiment can comprise temperature, humidity, wind speed and the like;
if it is detected that the feature attribute values corresponding to the child nodes respectively meet the preset conditions, the training node is set to be a leaf node, the child nodes with growth learning completed are set to be non-leaf nodes, and the leaf nodes and the non-leaf nodes are simultaneously input into the preset decision tree machine learning algorithm to train a corresponding target decision tree model, and further, the embodiment can identify corresponding suspicious error data through the target decision tree model in real time.
Specifically, in this embodiment, it should be noted that, in this embodiment, a corresponding man-machine interaction model is further constructed, and the man-machine interaction model constructed in this embodiment can accurately identify the interaction mode and the processing flow of the weather data service system, so that a process of performing real interface operations by a user, such as a data input process, a text clearing process, a popup window process, a filling form process, an element dragging process, a page switching process, a history record process, a cookie process, a mouse click process and other page interaction processes, can be simulated, and further, the artificial intelligent automatic processing of weather data is realized, and the processing efficiency of suspicious error data is correspondingly improved.
Further, in this embodiment, it should be noted that, the man-machine interaction model provided in this embodiment is implemented based on the existing MOOS (Maintenance Out Of Service) system, specifically, in this embodiment, page information data of the current MOOS system is first sorted out, and immediately, data analysis is performed on the page information data, and corresponding execution operation information is constructed, so that the establishment and use of the later man-machine interaction model are facilitated.
Based on this, the present embodiment further combines the existing intelligent man-machine interaction processing algorithm and the identification technology to perform secondary page analysis on the current MOOS system, and secondarily collates corresponding execution flow information, and further, the man-machine interaction model is constructed by combining the suspicious error data processing model provided in the above embodiment with the execution flow currently collated in real time.
Furthermore, in this embodiment, the suspicious error data processing model provided in the foregoing embodiment is further combined with the existing quality control system page operation model, that is, the suspicious error data processing model constructed in real time classifies the detected suspicious error data, and further combines with the execution flow corresponding to each type of suspicious error data, and meanwhile, the existing intelligent man-machine interaction processing and recognition technology is combined, so that the required man-machine interaction model for automatically processing the meteorological data can be finally constructed. Specifically, the automatic meteorological data processing man-machine interaction model provided by the embodiment is specifically composed of the existing deep convolutional neural network, knowledge graph reasoning, comprehensive reasoning, statistical reasoning, MIT AI Lab (namely the capability of autonomously browsing web pages to fill in blank knowledge), deep (the capability of solving the problem of requiring logic reasoning to complete by utilizing external memory) and other technologies. The weather data automatic processing man-machine interaction model provided by the embodiment can automatically match the image recognition, visual calculation and natural language processing results with corresponding features through the existing artificial intelligent algorithm, so that the weather data automatic processing man-machine interaction model has strong computing capacity and logic judgment capacity, and further can automatically perform corresponding logic judgment and business processing according to the page content recognized in real time.
In summary, according to the embodiment, the corresponding meteorological data automatic processing man-machine interaction model is constructed, so that the actual quality control interface operation of a user can be simulated in real time in the actual working process, and the corresponding meteorological data intelligent processing can be finally realized.
It should be noted that the foregoing implementation procedure is only for illustrating the feasibility of the present application, but this does not represent that the suspicious error data processing model training method of the present application has only one implementation procedure, and instead, the suspicious error data processing model training method of the present application can be implemented only by the method, which is included in the feasible embodiments of the present application.
In summary, the training method for the suspicious error data processing model provided by the embodiment of the invention can automatically process the identified suspicious error data through the trained decision tree model when the suspicious error data is identified by the meteorological data service system, and generate a corresponding suspicious error data processing result, thereby omitting the manual processing and analyzing process of the suspicious error data, automatically and comprehensively processing the suspicious error data, greatly improving the processing efficiency of the suspicious error data and being beneficial to improving the working efficiency of meteorological departments.
Referring to fig. 4, a training system for a suspicious error data processing model according to a sixth embodiment of the present invention is shown, where the system includes:
the acquisition module 12 is configured to acquire original historical suspicious error data, and perform classification processing on the original historical suspicious error data according to observation elements, so as to generate a plurality of corresponding feature data tables, where each feature data table includes a target value and a plurality of feature values;
the labeling module 22 is configured to sequentially add labeling values to the target value and the feature values to generate a corresponding target feature data table, and store the target feature data table into a target document;
the training module 32 is configured to train a preset decision tree machine learning algorithm based on the target document, and train a corresponding decision tree model, so that the decision tree model processes suspicious error data generated in real time.
In the above training system for suspicious error data processing model, the obtaining module 12 is specifically configured to:
when the original historical suspicious error data is obtained, detecting observation elements contained in the original historical suspicious error data, and splitting the observation elements to split the original historical suspicious error data into a plurality of corresponding observation element data sets, wherein the observation elements comprise air temperature, air pressure, wind direction, wind speed, relative humidity, evaporation capacity, sunlight, visibility and ground temperature;
And identifying a plurality of characteristic values contained in each observation element data set one by one, and predicting a corresponding target value according to the plurality of characteristic values so as to generate the characteristic data table according to the target value and the plurality of characteristic values.
In the above training system for suspicious error data processing model, the obtaining module 12 is further specifically configured to:
screening out feature information contained in a plurality of feature values in each observation element data set, wherein the feature information comprises quality control result description information, suspicious error type information corresponding to the quality control result description information and observation element information;
establishing a mapping relation among the quality control result description information, the suspected error type information and the observation element information, wherein the mapping relation has uniqueness, the suspected error type information comprises errors, suspicions and missing tests, and the observation element information comprises observation values;
and predicting a corresponding target processing mode in a preset conclusion database according to the mapping relation, and setting a processing result of the target processing mode as the target value.
In the above training system for suspicious error data processing model, the obtaining module 12 is further specifically configured to:
When the mapping relation is obtained, a target observation site corresponding to the current observation element data set is found out according to the mapping relation, and a target observation element of the current observation element data set in the target observation site is determined, wherein the target observation site comprises a national station and an regional station;
and predicting the corresponding target processing mode in the preset conclusion database according to the target observation element.
In the above suspicious error data processing model training system, the labeling module 22 is specifically configured to:
adding annotation columns at preset positions corresponding to the target value and the characteristic values one by one, and adding corresponding annotation values in the annotation columns one by one according to preset rules, wherein each target value and each annotation value corresponding to the characteristic value have uniqueness;
and respectively establishing the target value and the mapping relation between the characteristic value and the labeling value to generate the target characteristic data table.
In the above training system for suspicious error data processing model, the training module 32 is specifically configured to:
initializing the preset decision tree machine learning algorithm, inputting the target document into the initialized decision tree machine learning algorithm, and enabling the initialized decision tree machine learning algorithm to identify suspicious error information and suspicious error data circulation information in the target document;
And performing classification learning on the initialized decision tree machine learning algorithm through the suspicious error information and the suspicious error data circulation information so as to complete training of the preset decision tree machine learning algorithm.
In the above suspicious error data processing model training system, the suspicious error data processing model training system further includes a fusion module 42, where the fusion module 42 is specifically configured to:
and carrying out page analysis on the meteorological data service system through a man-machine interaction processing algorithm to sort out execution flow information in the meteorological data service system, and fusing the execution flow information into the decision tree model to generate a corresponding man-machine interaction model.
A seventh embodiment of the present invention provides a computer, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements the suspicious error data handling model training method provided in the above embodiment when executing the computer program.
An eighth embodiment of the present invention provides a storage medium having stored thereon a computer program which, when executed by a processor, implements the suspicious error data handling model training method provided by the above embodiment.
In summary, the training method, system, computer and storage medium for the suspicious error data processing model provided by the embodiment of the invention can automatically process the identified suspicious error data through the trained decision tree model and generate the corresponding suspicious error data processing result when the suspicious error data is identified by the meteorological data service system, thereby omitting the manual processing and analyzing process of the suspicious error data, automatically and comprehensively processing the suspicious error data, greatly improving the processing efficiency of the suspicious error data and being beneficial to improving the working efficiency of meteorological departments.
The above-described respective modules may be functional modules or program modules, and may be implemented by software or hardware. For modules implemented in hardware, the various modules described above may be located in the same processor; or the above modules may be located in different processors in any combination.
Logic and/or steps represented in the flowcharts or otherwise described herein, e.g., a ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). In addition, the computer readable medium may even be paper or other suitable medium on which the program is printed, as the program may be electronically captured, via, for instance, optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory.
It is to be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, may be implemented using any one or combination of the following techniques, as is well known in the art: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The foregoing examples illustrate only a few embodiments of the invention and are described in detail herein without thereby limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention. Accordingly, the scope of protection of the present invention is to be determined by the appended claims.
Claims (7)
1. A method for training a suspicious error data handling model, the method comprising:
Acquiring original historical suspicious error data, and classifying the original historical suspicious error data according to an observation element to generate a plurality of corresponding characteristic data tables, wherein each characteristic data table comprises a target value and a plurality of characteristic values;
adding labeling values to the target value and the characteristic values in sequence to generate a corresponding target characteristic data table, and storing the target characteristic data table into a target document;
training a preset decision tree machine learning algorithm based on the target document, and training a corresponding decision tree model so that the decision tree model processes suspicious error data generated in real time;
the step of classifying the original historical suspicious error data according to the observation elements to generate a plurality of corresponding characteristic data tables comprises the following steps:
when the original historical suspicious error data is obtained, detecting observation elements contained in the original historical suspicious error data, and splitting the observation elements to split the original historical suspicious error data into a plurality of corresponding observation element data sets, wherein each observation element corresponds to one observation element data set, and the observation elements comprise air temperature, air pressure, wind direction, wind speed, relative humidity, evaporation capacity, sunlight, visibility and ground temperature;
Identifying a plurality of characteristic values contained in each observation element data set one by one, and predicting a corresponding target value according to the plurality of characteristic values so as to generate the characteristic data table according to the target value and the plurality of characteristic values;
the step of identifying a plurality of characteristic values contained in each observation element data set one by one and predicting a corresponding target value according to the plurality of characteristic values comprises the following steps:
screening out feature information contained in a plurality of feature values in each observation element data set, wherein the feature information comprises quality control result description information, suspicious error type information corresponding to the quality control result description information and observation element information;
establishing a mapping relation among the quality control result description information, the suspected error type information and the observation element information, wherein the mapping relation has uniqueness, the suspected error type information comprises errors, suspicions and missing tests, and the observation element information comprises observation values;
predicting a corresponding target processing mode in a preset conclusion database according to the mapping relation, and setting a processing result of the target processing mode as the target value;
The step of predicting the corresponding target processing mode in the preset conclusion database according to the mapping relation comprises the following steps:
when the mapping relation is obtained, a target observation site corresponding to the current observation element data set is found out according to the mapping relation, and a target observation element of the current observation element data set in the target observation site is determined, wherein the target observation site comprises a national station and an regional station;
and predicting the corresponding target processing mode in the preset conclusion database according to the target observation element.
2. The suspicious error data handling model training method according to claim 1, wherein: the step of sequentially adding the labeling values to the target value and the feature values to generate a corresponding target feature data table comprises the following steps:
adding annotation columns at preset positions corresponding to the target value and the characteristic values one by one, and adding corresponding annotation values in the annotation columns one by one according to preset rules, wherein each target value and each annotation value corresponding to the characteristic value have uniqueness;
and respectively establishing the target value and the mapping relation between the characteristic value and the labeling value to generate the target characteristic data table.
3. The suspicious error data handling model training method according to claim 1, wherein: the step of training a preset decision tree machine learning algorithm based on the target document comprises the following steps:
initializing the preset decision tree machine learning algorithm, inputting the target document into the initialized decision tree machine learning algorithm, and enabling the initialized decision tree machine learning algorithm to identify suspicious error information and suspicious error data circulation information in the target document;
and performing classification learning on the initialized decision tree machine learning algorithm through the suspicious error information and the suspicious error data circulation information so as to complete training of the preset decision tree machine learning algorithm.
4. A suspected error data handling model training method according to claim 3, characterized in that: the step of performing classification learning on the initialized decision tree machine learning algorithm through the suspicious error information and the suspicious error data circulation information to complete training of the preset decision tree machine learning algorithm comprises the following steps:
when the suspicious error information and the suspicious error data circulation information are respectively acquired, generating a plurality of corresponding training samples according to the suspicious error information, and generating a plurality of corresponding training nodes according to the suspicious error data circulation information, wherein each training sample and each training node have uniqueness;
Converting each training sample into a plurality of corresponding feature vectors based on a CART algorithm, and inputting the plurality of feature vectors into a plurality of training nodes correspondingly respectively so as to perform corresponding growth learning on each training node and generate a plurality of corresponding child nodes;
performing continuous growth learning on the plurality of sub-nodes, and judging whether the characteristic attributes corresponding to the plurality of sub-nodes respectively meet preset conditions or not, wherein each sub-node corresponds to one type of characteristic attribute respectively;
if the fact that the characteristic attribute values corresponding to the child nodes respectively meet the preset conditions is detected, the training node is set to be a leaf node, the child nodes with the growth learning completed are set to be non-leaf nodes, and the leaf nodes and the non-leaf nodes are input into the preset decision tree machine learning algorithm at the same time, so that a corresponding target decision tree model is trained.
5. A suspected error data processing model training system for implementing a suspected error data processing model training method according to any of claims 1 to 4, the system comprising:
the acquisition module is used for acquiring original historical suspicious error data, classifying the original historical suspicious error data according to the observation factors to generate a plurality of corresponding characteristic data tables, wherein each characteristic data table comprises a target value and a plurality of characteristic values;
The labeling module is used for sequentially adding labeling values to the target value and the characteristic values to generate a corresponding target characteristic data table, and storing the target characteristic data table into a target document;
the training module is used for training a preset decision tree machine learning algorithm based on the target document and training a corresponding decision tree model so that the decision tree model can process suspicious error data generated in real time.
6. A computer comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the suspected error data handling model training method according to any of claims 1 to 4 when the computer program is executed by the processor.
7. A storage medium having stored thereon a computer program which when executed by a processor implements the suspected error data processing model training method according to any of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310425084.2A CN116150610B (en) | 2023-04-20 | 2023-04-20 | Training method, system, computer and storage medium for suspicious error data processing model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310425084.2A CN116150610B (en) | 2023-04-20 | 2023-04-20 | Training method, system, computer and storage medium for suspicious error data processing model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116150610A CN116150610A (en) | 2023-05-23 |
| CN116150610B true CN116150610B (en) | 2023-07-07 |
Family
ID=86358567
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310425084.2A Active CN116150610B (en) | 2023-04-20 | 2023-04-20 | Training method, system, computer and storage medium for suspicious error data processing model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116150610B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110458438A (en) * | 2019-08-05 | 2019-11-15 | 北京师范大学 | The calculation method and device of the impact factor of vegetation water use efficiency WUE |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10382842B2 (en) * | 2012-06-26 | 2019-08-13 | BTS Software Software Solutions, LLC | Realtime telemetry data compression system |
| US11610076B2 (en) * | 2019-08-07 | 2023-03-21 | Applied Materials, Inc. | Automatic and adaptive fault detection and classification limits |
| CN110705115B (en) * | 2019-10-10 | 2023-05-05 | 北京北科融智云计算科技有限公司 | Weather forecast method and system based on deep belief network |
| CN111045117B (en) * | 2020-01-13 | 2022-03-25 | 黑龙江省气候中心 | Climate monitoring and predicting platform |
| CN112464995A (en) * | 2020-11-06 | 2021-03-09 | 广东电网有限责任公司东莞供电局 | Power grid distribution transformer fault diagnosis method and system based on decision tree algorithm |
| CN114496243A (en) * | 2021-12-31 | 2022-05-13 | 东软集团股份有限公司 | Data processing method, data processing device, storage medium and electronic equipment |
-
2023
- 2023-04-20 CN CN202310425084.2A patent/CN116150610B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110458438A (en) * | 2019-08-05 | 2019-11-15 | 北京师范大学 | The calculation method and device of the impact factor of vegetation water use efficiency WUE |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116150610A (en) | 2023-05-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111507371B (en) | Method and device for automatically evaluating reliability of label on training image | |
| CN111401418A (en) | Employee dressing specification detection method based on improved Faster r-cnn | |
| CN112756759B (en) | Spot welding robot workstation fault judgment method | |
| CN112699953A (en) | Characteristic pyramid neural network architecture searching method based on multi-information path aggregation | |
| CN116416884B (en) | Testing device and testing method for display module | |
| CN115273112A (en) | Table identification method and device, electronic equipment and readable storage medium | |
| CN115034200A (en) | Drawing information extraction method and device, electronic equipment and storage medium | |
| CN116467674A (en) | Intelligent fault processing fusion updating system and method for power distribution network | |
| CN114491282B (en) | Abnormal user behavior analysis method and system based on cloud computing | |
| CN114818707A (en) | Automatic driving decision method and system based on knowledge graph | |
| CN114494168A (en) | Model determination, image recognition and industrial quality inspection method, equipment and storage medium | |
| CN116150610B (en) | Training method, system, computer and storage medium for suspicious error data processing model | |
| CN112182441A (en) | Method and device for detecting violation data | |
| CN115240145A (en) | Method and system for detecting illegal operation behaviors based on scene recognition | |
| CN115438190A (en) | Power distribution network fault decision-making assisting knowledge extraction method and system | |
| CN115619763A (en) | Power transmission line small part image target detection method based on multistage reasoning detection | |
| CN111144429A (en) | Object classification method and system, computer system and computer readable medium | |
| CN115204179A (en) | Entity relationship prediction method and device based on power grid public data model | |
| CN114708470A (en) | Illegal picture identification method, medium and computing device | |
| CN110458383B (en) | Method and device for realizing demand processing servitization, computer equipment and storage medium | |
| CN113689036A (en) | Thermal imager quality problem reason prediction method based on decision tree C4.5 algorithm | |
| CN109829500A (en) | A kind of position composition and automatic clustering method | |
| CN115952064B (en) | Multi-component fault interpretation method and device for distributed system | |
| CN113076454B (en) | Artificial intelligence-based element number analysis method and server | |
| CN112766387B (en) | Training data error correction method, device, equipment and storage medium |
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 |