CN107169593A - A kind of universities and colleges based on neutral net file line Forecasting Methodology - Google Patents
A kind of universities and colleges based on neutral net file line Forecasting Methodology Download PDFInfo
- Publication number
- CN107169593A CN107169593A CN201710273724.7A CN201710273724A CN107169593A CN 107169593 A CN107169593 A CN 107169593A CN 201710273724 A CN201710273724 A CN 201710273724A CN 107169593 A CN107169593 A CN 107169593A
- Authority
- CN
- China
- Prior art keywords
- universities
- colleges
- rule
- line
- year
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 21
- 230000007935 neutral effect Effects 0.000 title claims abstract description 20
- 238000012545 processing Methods 0.000 claims abstract description 5
- 238000012549 training Methods 0.000 claims description 26
- 210000002569 neuron Anatomy 0.000 claims description 14
- 238000003062 neural network model Methods 0.000 claims description 10
- 238000010606 normalization Methods 0.000 claims description 7
- 238000007405 data analysis Methods 0.000 abstract description 3
- 238000005516 engineering process Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
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
- G06Q50/10—Services
- G06Q50/20—Education
- G06Q50/205—Education administration or guidance
- G06Q50/2053—Education institution selection, admissions, or financial aid
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Educational Administration (AREA)
- Educational Technology (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Marketing (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Primary Health Care (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (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
Line Forecasting Methodology is filed the invention discloses a kind of universities and colleges based on neutral net, belongs to data analysis technique field, specifically includes following steps：Missing values and outlier processing are carried out to the data collected；Treated data are normalized；Data set is classified；Prediction is modeled respectively to the data set classified.This method can be filed to college entrance examination universities and colleges over the years point to be analyzed, so as to improve the accuracy rate that next year universities and colleges file prediction, be can be applied to universities and colleges and is filed line prediction.
Description
Technical field
The invention belongs to data mining and big data analysis technical field, it is related to the Predicting Technique that line is filed towards universities and colleges.
Background technology
The line of filing of each enrollment universities and colleges is the minimum control line (also known as saving control line) delimited in the provincial enrollment office of same batch
On the basis of, according to the school then this province enrollment plan, file ratio, examinee enter oneself for the examination the universities and colleges aspiration distribution situation and
What the achievement distribution situation of these examinees was determined jointly.It is provincial in the College Enrollment of each batch according to Ministry of Education's regulation
The ratio that examinee's archives are delivered by universities and colleges of enrollment office should be controlled 1：1~1：1.2 between.
And have every year many examinees be faced with the result finally enrolled higher position and it is low with regard to the problem of.So-called higher position is just
It is during first batch is enrolled, to be moved back shelves after filing and second lot can only be carried out due to the fraction of oneself not enough
Solicit parallel wish.It is so-called it is low be just exactly the achievement of oneself than the school entered oneself for the examination to file line high.
With the development of internet and big data technology, exploitation universities and colleges file line Predicting Technique, can effectively reduce student's chance
To higher position and low probability just, therefore, it is necessary to carry out brand-new research to this technology.
The content of the invention
Line Forecasting Methodology is filed it is an object of the invention to provide a kind of universities and colleges based on neutral net, to improve next year
Universities and colleges file the predictablity rate of line.
In order to solve the above technical problems, the technical solution adopted by the present invention is as follows.
A kind of universities and colleges based on neutral net file line Forecasting Methodology, it is characterised in that comprise the following steps：
Step one, missing values are carried out to the data collected and outlier is handled；
Treated data are normalized by step 2；
Step 3, classifies to data set；
Step 4, sets up neural network model to the data set classified and is predicted respectively.
The missing values and outlier processing include following four rule：
Rule 1.1, the poor missing values of the line of universities and colleges two is arranged to the poor average of two lines in remaining each year；
Rule 1.2, the poor outlier of two lines of universities and colleges is arranged to the average of remaining each line of year two difference；
Universities and colleges are filed line missing values and are arranged to year two line difference and save control line sum with this year by rule 1.3；
The normalization includes following two rules：
Rule 2.1, uses range method, the minimum value and maximum taken is that each line of year two is poor to the difference normalization of each line of year two
Minimum value and maximum；
Rule 2.2, range method is used to the normalization of each year precedence, and the minimum value and maximum taken is that each year precedence is minimum
Value and maximum.
It is described that data set is carried out to classify comprising three below rule：
Rule 3.1, it is training set 1 to filter out the universities and colleges that nearly 4 years two line differences are 0；
Rule 3.2, filters out nearly 4 years two lines difference and is respectively less than 5 points, and two lines difference is not 0 universities and colleges, is training set 2；
Rule 3.3, calculates 4 years precedence deviations of remaining universities and colleges, and deviation is training set 3 in 0~0.45 universities and colleges, and deviation exists
0.45~1 universities and colleges are training set 4, and the universities and colleges that deviation is more than 1 are training set 5, and the deviation is that annual precedence subtracts 4 years positions
Absolute value of the absolute value sum of secondary average than upper precedence average.
The neural network model of setting up includes following rule:
Rule 4.1, the two lines difference that the universities and colleges in training set 1 are given tacit consent to these universities and colleges' next years is 0 point；
Rule 4.2, to two lines poor historical data of the universities and colleges in training set 2 using the first four years, is carried out using neutral net
Modeling, wherein the data of the 4th year are input as remaining is exported.
Rule 4.3, the precedence historical data that the first four years are utilized respectively to training set 35 is modeled using neutral net, its
In the 4th year data as export remaining be input.
The neural network model is double-deck hidden layer, and the neuron of first layer hidden layer is not less than 20 neurons, the
Two layers of hidden layer are not less than 15 neurons.
The present invention has beneficial effect.Data analysis is to ensure that universities and colleges file line and predict accurate key technology.The present invention
A kind of data prediction technology proposed, is characterized in that the feature for taking into full account data is classified to data, by collecting
Data carry out the processing of missing values and outlier, treated data are normalized, data set is classified, it is right
The data set classified is modeled the technical methods such as prediction respectively, significantly improves the accuracy that universities and colleges file line prediction.
Embodiment
Technical scheme is described in further details with reference to embodiment.
With the data instance that 12 universities and colleges recruited student in Jiangsu Province are over the years, next year is carried out to this 12 universities and colleges and files line
Prediction, table 1 is the original table of this 12 universities and colleges' data over the years.
1 12 universities and colleges of table data original table over the years
By the information of table 1, after step one under this invention is handled, table 2 is obtained.
Table after the missing values of table 2. and outlier processing
By the information of table 2, after step 2 under this invention is handled, table 3 is obtained.
Table after the data normalization of table 3
By the information of table 3, after step 3 under this invention is handled, table 4~8 is obtained.
The training set 1 of table 4
The training set 2 of table 5.
The training set 3 of table 6.
The training set 4 of table 7.
The training set 5 of table 8.
The two lines difference of universities and colleges' next year in training set 1 is predicted as 0.
By the universities and colleges in training set 2 using 2012-2015 the poor historical data of two lines (2012-2014 is to input,
2015 be output) it is modeled using neutral net, by the use of 2013-2015 data as input, utilize the mould established
Type predicts that two lines of 2016 are poor.Wherein neural network model is double-deck hidden layer, and first layer hidden layer has 20 neurons,
Second hidden layer has 15 neurons.
By the universities and colleges in training set 3 using 2012-2015 precedence historical data (2012-2014 be input, 2015
For output) it is modeled using neutral net, it is pre- using the model established by the use of 2013-2015 data as input
The precedence of survey 2016.Wherein neural network model is double-deck hidden layer, and first layer hidden layer has 20 neurons, second
Hidden layer has 15 neurons.
By the universities and colleges in training set 4 using 2012-2015 precedence historical data (2012-2014 be input, 2015
For output) it is modeled using neutral net, it is pre- using the model established by the use of 2013-2015 data as input
The precedence of survey 2016.Wherein neural network model is double-deck hidden layer, and first layer hidden layer has 20 neurons, second
Hidden layer has 15 neurons.
By the universities and colleges in training set 5 using 2012-2015 precedence historical data (2012-2014 be input, 2015
For output) it is modeled using neutral net, it is pre- using the model established by the use of 2013-2015 data as input
The precedence of survey 2016.Wherein neural network model is double-deck hidden layer, and first layer hidden layer has 20 neurons, second
Hidden layer has 15 neurons.
On actual test collection, the test effect of model is as follows.
Accuracy：Error=0,24.79%
Error≤1,51.24%
Error≤2,65.7%
Error≤1,75.2%
And the test effect being modeled of not classified with two lines difference purely is as follows.
Accuracy：Error=0,21.49%
Error≤1,36.37%
Error≤2,54.55%
Error≤1,65.71%
As can be seen here, a kind of universities and colleges based on neutral net that the application is proposed file line Forecasting Methodology with higher standard
True rate.
Claims (6)
1. a kind of universities and colleges based on neutral net file line Forecasting Methodology, it is characterised in that comprise the following steps：
Step one, missing values are carried out to the data collected and outlier is handled；
Treated data are normalized by step 2；
Step 3, classifies to data set；
Step 4, sets up neural network model to the data set classified and is predicted respectively.
2. a kind of universities and colleges based on neutral net according to claim 1 file line Forecasting Methodology, it is characterised in that described
Missing values and outlier processing include following four rule：
Rule 1.1, the poor missing values of the line of universities and colleges two is arranged to the poor average of two lines in remaining each year；
Rule 1.2, the poor outlier of two lines of universities and colleges is arranged to the average of remaining each line of year two difference；
Universities and colleges are filed line missing values and are arranged to year two line difference and save control line sum with this year by rule 1.3.
3. a kind of universities and colleges based on neutral net according to claim 1 file line Forecasting Methodology, it is characterised in that described
Normalization includes following two rules：
Rule 2.1, range method is used to the difference normalization of each line of year two, and the minimum value and maximum taken is that each line of year two difference is minimum
Value and maximum；
Rule 2.2, to each year precedence normalization use range method, the minimum value and maximum taken be each year precedence minimum value and
Maximum.
4. a kind of universities and colleges based on neutral net according to claim 1 file line Forecasting Methodology, it is characterised in that described
Classification is carried out to data set and includes three below rule：
Rule 3.1, it is training set 1 to filter out the universities and colleges that nearly 4 years two line differences are 0；
Rule 3.2, filters out nearly 4 years two lines difference and is respectively less than 5 points, and two lines difference is not 0 universities and colleges, is training set 2；
Rule 3.3, calculates 4 years precedence deviations of remaining universities and colleges, and deviation is training set 3 in 0~0.45 universities and colleges, deviation 0.45~
1 universities and colleges are training set 4, and the universities and colleges that deviation is more than 1 are training set 5, and the deviation is that annual precedence subtracts 4 years precedence averages
Absolute value of the absolute value sum than upper precedence average.
5. a kind of universities and colleges based on neutral net according to claim 1-4 file line Forecasting Methodology, it is characterised in that institute
State and set up neural network model comprising following rule:
Rule 4.1, the two lines difference that the universities and colleges in training set 1 are given tacit consent to these universities and colleges' next years is 0 point；
Rule 4.2, to two lines poor historical data of the universities and colleges in training set 2 using the first four years, is modeled using neutral net,
The data of wherein the 4th year are input as remaining is exported.
Rule 4.3, the precedence historical data that the first four years are utilized respectively to training set 3~5 is modeled using neutral net, wherein
The data of the 4th year are input as remaining is exported.
6. a kind of universities and colleges based on neutral net according to claim 1 file line Forecasting Methodology, it is characterised in that described
Neural network model is double-deck hidden layer, and the neuron of first layer hidden layer is not less than 20 neurons, and second layer hidden layer is not
Less than 15 neurons.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710273724.7A CN107169593A (en) | 2017-04-25 | 2017-04-25 | A kind of universities and colleges based on neutral net file line Forecasting Methodology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710273724.7A CN107169593A (en) | 2017-04-25 | 2017-04-25 | A kind of universities and colleges based on neutral net file line Forecasting Methodology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107169593A true CN107169593A (en) | 2017-09-15 |
Family
ID=59812679
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710273724.7A Pending CN107169593A (en) | 2017-04-25 | 2017-04-25 | A kind of universities and colleges based on neutral net file line Forecasting Methodology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107169593A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108629454A (en) * | 2018-05-04 | 2018-10-09 | 上饶市普适科技有限公司 | The method for filing line with two or three recipe prediction colleges and universities |
| CN108764718A (en) * | 2018-05-28 | 2018-11-06 | 王春宁 | Selection method, system are estimated and volunteered to college entrance examination score based on deep learning algorithm |
-
2017
- 2017-04-25 CN CN201710273724.7A patent/CN107169593A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108629454A (en) * | 2018-05-04 | 2018-10-09 | 上饶市普适科技有限公司 | The method for filing line with two or three recipe prediction colleges and universities |
| CN108764718A (en) * | 2018-05-28 | 2018-11-06 | 王春宁 | Selection method, system are estimated and volunteered to college entrance examination score based on deep learning algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110070713B (en) | Traffic flow prediction method based on bidirectional nested LSTM neural network | |
| Tsamboulas et al. | Use of multicriteria methods for assessment of transport projects | |
| CN103745273B (en) | Semiconductor fabrication process multi-performance prediction method | |
| CN103632203B (en) | A kind of power distribution network division of the power supply area method based on overall merit | |
| CN106909933A (en) | A kind of stealing classification Forecasting Methodology of three stages various visual angles Fusion Features | |
| CN104408562B (en) | A kind of photovoltaic system generating efficiency comprehensive estimation method based on BP neural network | |
| CN107315884A (en) | A kind of building energy consumption modeling method based on linear regression | |
| CN110111024A (en) | Scientific and technological achievement market value evaluation method based on AHP fuzzy comprehensive evaluation model | |
| CN105956779A (en) | Power transformer running state evaluation method and power transformer running state evaluation device | |
| CN104537433A (en) | Sold electricity quantity prediction method based on inventory capacities and business expansion characteristics | |
| CN110047291A (en) | A kind of Short-time Traffic Flow Forecasting Methods considering diffusion process | |
| CN109063892A (en) | Industry watt-hour meter prediction technique based on BP-LSSVM combination optimization model | |
| CN107203752A (en) | A kind of combined depth study and the face identification method of the norm constraint of feature two | |
| CN105956768A (en) | Power generation enterprise competitiveness evaluation method based on combined weight determining and improved TOPSIS | |
| CN103632306A (en) | Distribution network power supply area division method based on clustering analysis | |
| CN106803010A (en) | For the Fuzzy Grey comprehensive evaluation method and device of low permeability reservoir quantitative assessment | |
| CN108415393A (en) | A kind of GaAs product quality consistency control method and system | |
| CN109299434B (en) | Cargo customs clearance big data is intelligently graded and sampling observation rate computing system | |
| CN103700030A (en) | Grey rough set-based power grid construction project post-evaluation index weight assignment method | |
| CN111626559A (en) | Main factor analysis method-based medium-voltage distribution network line loss key characteristic index extraction method and system | |
| CN110059765A (en) | A kind of mineral intelligent recognition categorizing system and method | |
| CN107169593A (en) | A kind of universities and colleges based on neutral net file line Forecasting Methodology | |
| CN108229011A (en) | A kind of shale lithofacies development Dominated Factors judgment method, equipment and storage device | |
| CN107403391A (en) | A kind of optimal screening method for oil field development pre-arranged project | |
| CN116109212B (en) | Airport operation efficiency evaluation index design and monitoring method |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170915 |