CN105512762A - Game numerical value launching estimation method and device based on correlation analysis - Google Patents
Game numerical value launching estimation method and device based on correlation analysis Download PDFInfo
- Publication number
- CN105512762A CN105512762A CN201510888956.4A CN201510888956A CN105512762A CN 105512762 A CN105512762 A CN 105512762A CN 201510888956 A CN201510888956 A CN 201510888956A CN 105512762 A CN105512762 A CN 105512762A
- Authority
- CN
- China
- Prior art keywords
- numerical value
- game
- index
- throws
- sample cycle
- 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
- 238000010219 correlation analysis Methods 0.000 title claims abstract description 48
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000012216 screening Methods 0.000 claims description 16
- 230000001419 dependent effect Effects 0.000 claims description 12
- 238000013480 data collection Methods 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 4
- 238000013461 design Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000003203 everyday effect Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000002354 daily effect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 238000012360 testing method 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
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Primary Health Care (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
The invention discloses a game numerical value launching estimation method and device based on correlation analysis. The method comprises the following steps: S1, the historical data of a plurality of game operation indexes and game numerical value launch indexes is collected; S2, the game operation indexes and the game numerical value launch indexes required by the numerical value launching estimation are screened; S3, based on the indexes determined in S2, adopting the historical data collected in S1, a sample period and a prediction period of the game numerical value launching are reasonably predicted and are determined through a typical correlation analysis method; S4, based on the indexes determined in S2, adopting historical data of a sample period, a matching linear coefficient of the game operation indexes and the game numerical value launch indexes is determined through the typical correlation analysis method, and the game operation data which will be bought in the prediction period next to the sample period is calculated based on the corresponding linear relation. According to the invention, the method can use the game historical data to monitor the safety and rationality of the numerical launching.
Description
Technical field
The present invention relates to a kind of game numerical value based on correlation analysis and throw in appraisal procedure and device.
Background technology
Game is a virtual world, player's essential income is in gaming numerical value, brought by numerical value that player is a series of to be experienced (enjoyment, sense of accomplishment, sense of defeat etc.) in gaming, numerical value in game is more, comprise experience, game money, ingot etc., in current game numerical value input and reclaim by the input of numerical value sponsor person according to the self-designed model amount of carrying out, product is main by virtue of experience to the input of the activity amount of carrying out of different times, and actual play injected volume may be inconsistent with the expection of product.Meanwhile, in game, may there is affecting the improper game behavior of numerical value balance in game, as operating room's brush amount makes the large numerical quantity of output but consumption does not really increase, bug causes numerical value output to increase but active not increase.
But, the amount that logarithm value is not thrown at present is effectively assessed, how much how many experiences of such as having thrown in is enlivened account or increases how many consumption if should being brought, the amount that current sponsor's logarithm value is thrown in also is only the state of moving steadily, and is also delayed to the monitoring of operating room.
Summary of the invention
Fundamental purpose of the present invention is to overcome the deficiencies in the prior art, provides a kind of game numerical value based on correlation analysis to throw in appraisal procedure and device, solves online game numerical value and throws in rationality and complete problem.
For achieving the above object, the present invention is by the following technical solutions:
Game numerical value based on correlation analysis throws in an appraisal procedure, comprises the following steps:
S1, collect the historical data that multinomial game operation indicator and multinomial game numerical value throw in index;
S2, screening game numerical value throw in the game operation indicator needed for assessing and game numerical value throws in index;
S3, the index determined based on step S2, utilize the historical data that step S1 collects, and uses the method for canonical correlation analysis to determine to throw in game numerical value sample cycle and predetermined period of carrying out reasonable prediction;
S4, the index determined based on step S2, utilize the historical data of a sample cycle, game numerical value throws in the linear coefficient of index and game operation indicator to use the method for canonical correlation analysis to determine, and calculates the game operation data will brought in predetermined period of this sample cycle adjacent based on corresponding linear relationship.
Further:
Described method is further comprising the steps of:
S5, by step S4 predict game operation data with actual play operation data compared with, thus judge play numerical value input whether reasonable.
Step S2 comprises: judging that the game operation indicator of playing needed for numerical value input assessment is thrown in index with game numerical value and whether had predefined strong correlation, only retaining one of them index for there being multiple indexs of strong correlation.
Step S2 comprises:
Using each game operation indicator as one group of dependent variable Y, index is thrown in as one group of independent variable X using each game numerical value, set up the linear combination U be made up of described one group of dependent variable Y and the linear combination V be made up of described one group of independent variable X respectively, and to make the related coefficient between linear combination U and linear combination V reach maximal value for condition, determine the coefficient value of variable X separately in the coefficient value of each dependent variable Y in linear combination U and linear combination V;
Judge whether the coefficient value gap of each index in linear combination U and linear combination V exceedes predetermined threshold, if exceed predetermined threshold, from linear combination U and linear combination V, reject corresponding index, otherwise retain corresponding index.
The coefficient value of all indexs retained is within same number of stages.
Step S3 comprises:
Using historical data as training set, select initial predicted cycle and sample cycle;
Utilize the linear fit that the historical data of described sample cycle is carried out based on canonical correlation analysis, and determine the maximum distance between sample value in described sample cycle and fitting a straight line, as standard critical distance;
If the predicted value determined based on the relation of described linear fit in described predetermined period exceedes described standard critical distance to the distance of described fitting a straight line, then adjust sample cycle and/or predetermined period, until the predicted value in described predetermined period is no more than described standard critical distance to the distance of described fitting a straight line.
Wherein said predetermined period time is adjacent described sample cycle above.
Game numerical value based on correlation analysis throws in an apparatus for evaluating, comprising:
Data collection module, it collects the historical data of multinomial game operation indicator and multinomial game numerical value input index;
Index screening module, it determines that the game operation indicator of playing needed for numerical value input assessment throws in index with game numerical value;
Period determination module, its index determined based on described index screening module, utilizes the historical data that described data collection module is collected, and uses the method for canonical correlation analysis to determine to throw in game numerical value sample cycle and predetermined period of carrying out reasonable prediction;
Prediction module, its index determined based on described index screening module, utilize the historical data of a sample cycle, game numerical value throws in the linear coefficient of index and game operation indicator to use the method for canonical correlation analysis to determine, and calculates based on the linear relationship of correspondence the game operation data will brought in predetermined period corresponding with this sample cycle.
Further:
This device also comprises analysis module, its by the game operation data of prediction compared with actual play operation data, thus judge game numerical value throw in whether reasonable.
Described period determination module comprises:
Initial period selects module, using historical data as training set, selects initial predicted cycle and sample cycle;
Standard critical distance determination module, utilizes the linear fit that the historical data of described sample cycle is carried out based on canonical correlation analysis, determines the maximum distance between sample value in described sample cycle and fitting a straight line, as standard critical distance;
Period modulation module, if the predicted value determined based on the relation of described linear fit in described predetermined period exceedes described standard critical distance to the distance of described fitting a straight line, then adjust sample cycle and/or predetermined period, until the predicted value in described predetermined period is no more than described standard critical distance to the distance of described fitting a straight line.
Beneficial effect of the present invention:
The present invention proposes a kind of numerical value based on correlation analysis and throws in appraisal procedure, the present invention is based on the method for canonical correlation analysis, utilize game history data to monitor security and the rationality of numerical value input, wherein, by data collection step, collect game operation and numerical value input index of correlation; By screening index step, determine the index needed for assessing; By sample cycle and predetermined period determining step, determine as the rational sample time of game and predicted time cycle; By prediction steps, according to the determined cycle, throw in situation according to historical data prediction future values.When unexpected output a large amount of numerical value (as bug causes, runs input, operating room's brush amount) of playing, but when operation indicator does not obtain the lifting of respective degrees, distant just with actual operation indicator of the operation indicator utilizing canonical correlation analysis to predict, utilize the early warning that method of the present invention then can provide numerical value to throw in product, the design effort of auxiliary game sponsor.The present invention can be used for solving online game numerical value and throws in rationality and complete problem.Owing to throwing in situation according to historical data prediction future values, the present invention can help sponsor understand the same day numerical value throw in compare history throw in whether have larger fluctuation, can judge compared with actual operation data simultaneously numerical value throw in the operation data brought change whether reasonable, if occur abnormal, can instead push away is which kind of numerical value is thrown in unreasonable, to locate the reason of generation problem.Such as, the linear coefficient that numerical value throws in index and operation indicator has been simulated by stable historical data, show that numerical value throws in the predicted value of index by linear relationship, compare the distance of actual value and predicted value, if actual value departs from predicted value comparatively far, need special concern.
Further, the present invention proposes the method in screening index and predicted time cycle, is more conducive to utilizing large data, in real time reliably for sponsor provides assessment and the early warning of numerical value quantum of output.
Accompanying drawing explanation
Fig. 1 is the process flow diagram of the game numerical value input appraisal procedure embodiment that the present invention is based on correlation analysis.
Embodiment
Below embodiments of the present invention are elaborated.It is emphasized that following explanation is only exemplary, instead of in order to limit the scope of the invention and apply.
Consult Fig. 1, in one embodiment, a kind of game numerical value based on correlation analysis throws in appraisal procedure, comprises the following steps:
S1, collect the historical data that multinomial game operation indicator and multinomial game numerical value throw in index;
S2, screening game numerical value throw in the game operation indicator needed for assessing and game numerical value throws in index;
S3, the index determined based on step S2, utilize the historical data that step S1 collects, and uses the method for canonical correlation analysis to determine to throw in game numerical value sample cycle and predetermined period of carrying out reasonable prediction;
S4, the index determined based on step S2, utilize the historical data of a sample cycle, game numerical value throws in the linear coefficient of index and game operation indicator to use the method for canonical correlation analysis to determine, and calculates the game operation data will brought in predetermined period of this sample cycle adjacent based on corresponding linear relationship.
In a preferred embodiment, described method is further comprising the steps of: S5, by step S4 predict game operation data with actual play operation data compared with, thus judge play numerical value input whether reasonable.
According to said method, rationality and completeness can be thrown in by Efficient Evaluation online game numerical value.This method is based on Canonical Correlation Analysis, the linear coefficient that numerical value throws in index and operation indicator is simulated by stable historical data, show that numerical value throws in the predicted value of index by linear relationship, the relatively distance of actual value and predicted value, if actual value departs from predicted value comparatively far, needs special concern.By throwing in situation according to historical data prediction future values; this method can help sponsor understand the same day numerical value throw in compare history throw in whether have larger fluctuation; can judge compared with actual operation data simultaneously numerical value throw in the operation data brought change whether reasonable; if occur, abnormal can also analysis is which kind of numerical value is thrown in unreasonable, so that the reason that orientation problem occurs.Such as, when unexpected output a large amount of numerical value (as bug causes, runs input, operating room's brush amount) of playing, but when operation indicator does not obtain the lifting of respective degrees, distant just with actual operation indicator of the operation indicator utilizing canonical correlation analysis to predict, therefore, utilize the early warning that method of the present invention then can provide numerical value to throw in product, the design effort of auxiliary game sponsor.
In a preferred embodiment, step S2 can comprise: judging that the game operation indicator of playing needed for numerical value input assessment is thrown in index with game numerical value and whether had predefined strong correlation, only retaining one of them index for there being multiple indexs of strong correlation.
In a preferred embodiment, the coefficient weights in canonical correlation analysis can be relied on to reject the index of strong correlation, and specifically, step S2 comprises:
Using each game operation indicator as one group of dependent variable Y, index is thrown in as one group of independent variable X using each game numerical value, set up the linear combination U be made up of described one group of dependent variable Y and the linear combination V be made up of described one group of independent variable X respectively, and to make the related coefficient between linear combination U and linear combination V reach maximal value for condition, determine the coefficient value of variable X separately in the coefficient value of each dependent variable Y in linear combination U and linear combination V;
Judge whether the coefficient value gap of each index in linear combination U and linear combination V exceedes predetermined threshold, if exceed predetermined threshold, from linear combination U and linear combination V, reject corresponding index, otherwise retain corresponding index.
The coefficient value (i.e. the coefficient weights of index) of all indexs retained is within same number of stages.
In a preferred embodiment, step S3 comprises:
Using historical data as training set, select initial predicted cycle and sample cycle;
Utilize the linear fit that the historical data of described sample cycle is carried out based on canonical correlation analysis, and determine the maximum distance between sample value in described sample cycle and fitting a straight line, as standard critical distance;
If the predicted value determined based on the relation of described linear fit in described predetermined period exceedes described standard critical distance to the distance of described fitting a straight line, then adjust sample cycle and/or predetermined period, until the predicted value in described predetermined period is no more than described standard critical distance to the distance of described fitting a straight line.
Wherein said predetermined period time is adjacent described sample cycle above.
Game numerical value based on correlation analysis throws in an apparatus for evaluating, comprising:
Data collection module, it collects the historical data of multinomial game operation indicator and multinomial game numerical value input index;
Index screening module, it determines that the game operation indicator of playing needed for numerical value input assessment throws in index with game numerical value;
Period determination module, its index determined based on described index screening module, utilizes the historical data that described data collection module is collected, and uses the method for canonical correlation analysis to determine to throw in game numerical value sample cycle and predetermined period of carrying out reasonable prediction;
Prediction module, its index determined based on described index screening module, utilize the historical data of a sample cycle, game numerical value throws in the linear coefficient of index and game operation indicator to use the method for canonical correlation analysis to determine, and calculates based on the linear relationship of correspondence the game operation data will brought in predetermined period corresponding with this sample cycle.
In a preferred embodiment, this device also comprises analysis module, its by the game operation data of prediction compared with actual play operation data, thus judge game numerical value throw in whether reasonable.
In a preferred embodiment, described period determination module comprises:
Initial period selects module, using historical data as training set, selects initial predicted cycle and sample cycle;
Standard critical distance determination module, utilizes the linear fit that the historical data of described sample cycle is carried out based on canonical correlation analysis, determines the maximum distance between sample value in described sample cycle and fitting a straight line, as standard critical distance;
Period modulation module, if the predicted value determined based on the relation of described linear fit in described predetermined period exceedes described standard critical distance to the distance of described fitting a straight line, then adjust sample cycle and/or predetermined period, until the predicted value in described predetermined period is no more than described standard critical distance to the distance of described fitting a straight line.
Be described below by way of embodiment more specifically.
1, Data Collection: collect and store game operation and numerical value input index of correlation
First, it (can be numerical value index of correlation actual in game that the numerical value that can calculate in game recent every day throws in index, as experience output, ingot output, ingot recovery etc.) and operation indicator (can be game popularity and business revenue index of correlation, as logged in role's number, paying volume etc.).The numerical value of different game throws in index and operation indicator may there is some difference, the numerical value of the feature statistical correlation of each game can be relied on to throw in index and operation indicator, in order to avoid omit important indicator.
The index storage mode that every day calculates can as following table 1.Data Collection memory technology can by the distributed data processing framework of Hadoop+Hive.
Table 1 index storage list
2, screening index: determine the index predicted
The part index number that can will reject in strong correlation index, only retains one of them index.Reject strong correlation index, can avoid causing the weight of other indexs too low because of the existence of two or more strong correlation index, and the object of the overall monitor of indices cannot be reached.The correlativity of the method test coefficient of canonical correlation analysis can be utilized, reject strong correlation index in one.In first group of index as certain game canonical correlation analysis, login role number (dau_acc) in operation indicator and numerical value ingot acquisition role's number (yuanbaogain_cnt) thrown in index is that the index of a pair strong correlation is (because analyzed game exists the system logging in and can get ingot, then there is strong correlation in login role's number and ingot acquisition role number), two indices exists simultaneously, can have influence on the weight of other indexs, then can reject one of them index
Can use canonical correlation analysis, using operation indicator as dependent variable, in game, numerical value throws in index as independent variable, analyzes and screen index.Canonical correlation analysis technology is that the one of multiple regression and correlation analysis extends, and needs the linear relationship between investigation one group of independent variable X and group dependent variable Y (two or several Y variable).
Need to do following judgement in analytic process:
1) according to canonical correlation analysis, first a linear combination of this group dependent variable Y is found, as U1=a1Y1+a2Y2...aQYQ, and a linear combination of this group independent variable X, as V1=b1X1+b2X2++bPXP, wherein optional one group of coefficient (coefficient sets a and coefficient sets b), can calculate the value of U1 and V1 respectively with each sample point.Utilize the N number of sample point comprised in sample, just can calculate the N number of several right of U1 and V1, then calculate the simple correlation coefficient of U1 and V1.Coefficient sets a selected before the size of the related coefficient calculated depends on and coefficient sets b.According to the present embodiment, the coefficient sets a chosen and the value of coefficient sets b make the related coefficient between U1 and V1 reach maximal value.Select coefficient sets a and coefficient sets b like this, the linear combination U1 obtained is called the first canonical variable of variable Y, and V1 is called the first canonical variable of variable X, and it makes there is maximum linear correlation between X variables collection and Y variables collection.
2) judge that whether the gap of the coefficient value (coefficient weights) of each index in the first canonical variable is excessive.The coefficient of the first canonical variable is calculated by statistical software, specifically by the coefficient of the first canonical variable of the cancor function calculating in R software, in coefficient, the coefficient of dau_acc and yuanbaogain_cnt is tens times of the coefficient of other indexs even hundred times, the weight of this two indices is excessive like this, the contribution of other indexs to linear fit can be made too small, therefore index excessive for coefficient is rejected.
In a preferred embodiment, ensure that the coefficient weights of all indexs is within same number of stages as far as possible.
3, the determination of sample cycle and predetermined period
Stable game history data can be utilized to determine sample cycle and predetermined period, first enumerate predetermined period, observe the prediction of stablizing data and whether be zone of reasonableness, thus determine sample cycle and predetermined period.Because hand trip game environment pace of change is very fast, the game environment of this month may be different with last month, therefore first determines that the rational time cycle is predicted according to this cycle again.Within the cycle of known games ambient stable, enumerate sample cycle and predetermined period, such as certain game is that a numerical value throws in index and operation indicator all reasonably time period within September 1 to these 30 days September 30, enumerate during this period of time in rational sample time and predicted time (ensure predicted time section adjacent with sample time section, and ensure that sample cycle least unit is 7 days due to the relation of self actual cycle of playing), such as following table:
| Sample time section | Predicted time section |
| 1-September 7 September | September 8 |
| 1-September 14 September | September 15 |
| …… |
Verify different sample cycle and expection cycle one by one, if the value predicted by canonical correlation analysis of selected sample cycle and actual value gap all less, then sample cycle and predetermined period reasonable.
Can by the matching of canonical correlation analysis, judge whether in the reasonable scope predicted value, the straight line of the first canonical variable matching of canonical correlation analysis is aY=bX, because the numerical indication in sample date in historical data and forecast date and operation indicator are all rational, then predicted value should with the close together of straight line aY=bX, if sample cycle and predetermined period long or too short all likely make predicted value and air line distance far away, judge that standard far away is: with the distance of the solstics of straight line to fitting a straight line in sample, the departure degree of predicted value can not exceed the departure degree farthest in sample, if the predicted value in predetermined period is above standard to the distance of straight line, then need adjustment sample cycle or predetermined period, circulation like this can search out rational sample cycle and predetermined period.Such as, can the sample point of 4 weeks (sample cycle) prediction following 2 days (predetermined period) before cycle calculations and future position, pass sample cycle and predetermined period on a timeline, if the prediction finding part-time point is rational, but the prediction of part-time point is irrational, namely there is future position to be above standard critical distance to the distance of fitting a straight line, then need to continue change sample cycle and predetermined period, until predicted value is all in the zone of reasonableness being no more than the departure degree farthest in sample.
4, predict and analyze numerical value throw in whether reasonable
Sample cycle and predetermined period just can start prediction after determining.Such as, determining sample cycle by the data of certain game is 2 weeks, predetermined period is 3 days, suppose that today is October 15, the historical data in 1-October 14 October can be used to calculate the first canonical variable coefficient a and b of canonical correlation analysis, then usage factor daily calculates the predicted value in October 15, October 16, three days on the 17th October successively, compare with the criterion distance of sample cycle again, the term of validity due to sample coefficient is 3 days, predict again after then recalculating coefficient a and b after October 17, the like.
Index is thrown in by the numerical value predicting new one day, the gap of predicted value and actual value can also be judged further, the rationality that new one day numerical value is thrown in can being assessed, whether rationally throwing in the operation data change that brings for monitoring numerical value, thus the design effort of auxiliary game sponsor.
Above content combines concrete/preferred embodiment further description made for the present invention, can not assert that specific embodiment of the invention is confined to these explanations.For general technical staff of the technical field of the invention; without departing from the inventive concept of the premise; its embodiment that can also describe these makes some substituting or modification, and these substitute or variant all should be considered as belonging to protection scope of the present invention.
Claims (10)
1. the game numerical value based on correlation analysis throws in an appraisal procedure, it is characterized in that, comprises the following steps:
S1, collect the historical data that multinomial game operation indicator and multinomial game numerical value throw in index;
S2, screening game numerical value throw in the game operation indicator needed for assessing and game numerical value throws in index;
S3, the index filtered out based on step S2, utilize the historical data that step S1 collects, and uses the method for canonical correlation analysis to determine to throw in game numerical value sample cycle and predetermined period of carrying out reasonable prediction;
S4, the index filtered out based on step S2, utilize the historical data of a sample cycle, game numerical value throws in the linear coefficient of index and game operation indicator to use the method for canonical correlation analysis to determine, and calculates the game operation data will brought in predetermined period of this sample cycle adjacent based on corresponding linear relationship.
2. game numerical value as claimed in claim 1 throws in appraisal procedure, it is characterized in that, further comprising the steps of:
S5, by step S4 predict game operation data with actual play operation data compared with, thus judge play numerical value input whether reasonable.
3. game numerical value as claimed in claim 1 throws in appraisal procedure, it is characterized in that, step S2 comprises: judging that the game operation indicator of playing needed for numerical value input assessment is thrown in index with game numerical value and whether had predefined strong correlation, only retaining one of them index for there being multiple indexs of strong correlation.
4. game numerical value as claimed in claim 1 throws in appraisal procedure, and it is characterized in that, step S2 comprises:
Using each game operation indicator as one group of dependent variable Y, index is thrown in as one group of independent variable X using each game numerical value, set up the linear combination U be made up of described one group of dependent variable Y and the linear combination V be made up of described one group of independent variable X respectively, and to make the related coefficient between linear combination U and linear combination V reach maximal value for condition, determine the coefficient value of variable X separately in the coefficient value of each dependent variable Y in linear combination U and linear combination V;
Judge whether the coefficient value gap of each index in linear combination U and linear combination V exceedes predetermined threshold, if exceed predetermined threshold, from linear combination U and linear combination V, reject corresponding index, otherwise retain corresponding index.
5. numerical value of playing as claimed in claim 4 throws in appraisal procedure, and it is characterized in that, the coefficient value of all indexs of reservation is within same number of stages.
6. numerical value of playing as described in any one of claim 1 to 5 throws in appraisal procedure, and it is characterized in that, step S3 comprises:
Using historical data as training set, select initial predicted cycle and sample cycle;
Utilize the linear fit that the historical data of described sample cycle is carried out based on canonical correlation analysis, and determine the maximum distance between sample value in described sample cycle and fitting a straight line, as standard critical distance;
If the predicted value determined based on the relation of described linear fit in described predetermined period exceedes described standard critical distance to the distance of described fitting a straight line, then adjust sample cycle and/or predetermined period, until the predicted value in described predetermined period is no more than described standard critical distance to the distance of described fitting a straight line.
7. numerical value of playing as claimed in claim 6 throws in appraisal procedure, and it is characterized in that, wherein said predetermined period time is adjacent described sample cycle above.
8. the game numerical value based on correlation analysis throws in an apparatus for evaluating, it is characterized in that, comprising:
Data collection module, it collects the historical data of multinomial game operation indicator and multinomial game numerical value input index;
Index screening module, the game operation indicator that its screening game numerical value is thrown in needed for assessment throws in index with game numerical value;
Period determination module, its index filtered out based on described index screening module, utilizes the historical data that described data collection module is collected, and uses the method for canonical correlation analysis to determine to throw in game numerical value sample cycle and predetermined period of carrying out reasonable prediction;
Prediction module, its index filtered out based on described index screening module, utilize the historical data of a sample cycle, game numerical value throws in the linear coefficient of index and game operation indicator to use the method for canonical correlation analysis to determine, and calculates based on the linear relationship of correspondence the game operation data will brought in predetermined period corresponding with this sample cycle.
9. game numerical value as claimed in claim 1 throws in apparatus for evaluating, it is characterized in that, also comprises analysis module, and it is by the game operation data of prediction with actual play operation data compared with, thus whether the numerical value input that judges to play is reasonable.
10. numerical value of playing as described in claim 8 or 9 throws in apparatus for evaluating, and it is characterized in that, described period determination module comprises:
Initial period selects module, using historical data as training set, selects initial predicted cycle and sample cycle;
Standard critical distance determination module, utilizes the linear fit that the historical data of described sample cycle is carried out based on canonical correlation analysis, determines the maximum distance between sample value in described sample cycle and fitting a straight line, as standard critical distance;
Period modulation module, if the predicted value determined based on the relation of described linear fit in described predetermined period exceedes described standard critical distance to the distance of described fitting a straight line, then adjust sample cycle and/or predetermined period, until the predicted value in described predetermined period is no more than described standard critical distance to the distance of described fitting a straight line.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510888956.4A CN105512762A (en) | 2015-12-04 | 2015-12-04 | Game numerical value launching estimation method and device based on correlation analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510888956.4A CN105512762A (en) | 2015-12-04 | 2015-12-04 | Game numerical value launching estimation method and device based on correlation analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105512762A true CN105512762A (en) | 2016-04-20 |
Family
ID=55720724
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510888956.4A Pending CN105512762A (en) | 2015-12-04 | 2015-12-04 | Game numerical value launching estimation method and device based on correlation analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105512762A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108854075A (en) * | 2018-05-30 | 2018-11-23 | 腾讯科技（深圳）有限公司 | Game role anomaly detection method and device |
| CN109857550A (en) * | 2019-01-07 | 2019-06-07 | 平安科技（深圳）有限公司 | Resource allocation method, device, equipment and storage medium based on machine learning |
| CN111224830A (en) * | 2018-11-23 | 2020-06-02 | 中国电信股份有限公司 | Data monitoring method and device, Internet of things network element and computer readable storage medium |
| CN112270436A (en) * | 2020-10-26 | 2021-01-26 | 北京明略昭辉科技有限公司 | Resource delivery effect evaluation method, device and system |
| CN113191810A (en) * | 2021-04-30 | 2021-07-30 | 网易（杭州）网络有限公司 | Game index prediction method and device and electronic equipment |
| CN113633992A (en) * | 2021-08-09 | 2021-11-12 | 网易（杭州）网络有限公司 | Game operation data prediction method and device, electronic equipment and storage medium |
| CN116663744A (en) * | 2023-06-20 | 2023-08-29 | 宁夏中昊银晨能源技术服务有限公司 | Energy consumption prediction method and system for near-zero energy consumption building |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110037707A1 (en) * | 2009-08-17 | 2011-02-17 | Nokia Corporation | Apparatus comprising an optically transparent sheet and related methods |
| CN102567536A (en) * | 2011-12-31 | 2012-07-11 | 南京柏梭信息科技有限公司 | Key performance target analyzing method based on data statistics |
| CN104346372A (en) * | 2013-07-31 | 2015-02-11 | 国际商业机器公司 | Method and device for evaluating prediction models |
| CN104899405A (en) * | 2014-03-04 | 2015-09-09 | 携程计算机技术（上海）有限公司 | Data prediction method and system and alarming method and system |
-
2015
- 2015-12-04 CN CN201510888956.4A patent/CN105512762A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110037707A1 (en) * | 2009-08-17 | 2011-02-17 | Nokia Corporation | Apparatus comprising an optically transparent sheet and related methods |
| CN102567536A (en) * | 2011-12-31 | 2012-07-11 | 南京柏梭信息科技有限公司 | Key performance target analyzing method based on data statistics |
| CN104346372A (en) * | 2013-07-31 | 2015-02-11 | 国际商业机器公司 | Method and device for evaluating prediction models |
| CN104899405A (en) * | 2014-03-04 | 2015-09-09 | 携程计算机技术（上海）有限公司 | Data prediction method and system and alarming method and system |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108854075A (en) * | 2018-05-30 | 2018-11-23 | 腾讯科技（深圳）有限公司 | Game role anomaly detection method and device |
| CN111224830A (en) * | 2018-11-23 | 2020-06-02 | 中国电信股份有限公司 | Data monitoring method and device, Internet of things network element and computer readable storage medium |
| CN109857550A (en) * | 2019-01-07 | 2019-06-07 | 平安科技（深圳）有限公司 | Resource allocation method, device, equipment and storage medium based on machine learning |
| CN112270436A (en) * | 2020-10-26 | 2021-01-26 | 北京明略昭辉科技有限公司 | Resource delivery effect evaluation method, device and system |
| CN113191810A (en) * | 2021-04-30 | 2021-07-30 | 网易（杭州）网络有限公司 | Game index prediction method and device and electronic equipment |
| CN113633992A (en) * | 2021-08-09 | 2021-11-12 | 网易（杭州）网络有限公司 | Game operation data prediction method and device, electronic equipment and storage medium |
| CN116663744A (en) * | 2023-06-20 | 2023-08-29 | 宁夏中昊银晨能源技术服务有限公司 | Energy consumption prediction method and system for near-zero energy consumption building |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105512762A (en) | Game numerical value launching estimation method and device based on correlation analysis | |
| Carey et al. | Predictive modelling of training loads and injury in Australian football | |
| Towlson et al. | When does the influence of maturation on anthropometric and physical fitness characteristics increase and subside? | |
| Varley et al. | Physical and technical performance of elite youth soccer players during international tournaments: influence of playing position and team success and opponent quality | |
| Yi et al. | Technical and physical match performance of teams in the 2018 FIFA World Cup: Effects of two different playing styles | |
| Li et al. | Is international junior success a reliable predictor for international senior success in elite combat sports? | |
| Bradley et al. | The influence of situational variables on ball possession in the English Premier League | |
| CN110417721A (en) | Safety risk estimating method, device, equipment and computer readable storage medium | |
| Young et al. | Movement demands in Australian rules football as indicators of muscle damage | |
| Berri et al. | Catching a draft: On the process of selecting quarterbacks in the National Football League amateur draft | |
| CN104954342B (en) | A kind of safety evaluation method and device | |
| Barr et al. | Validity and reliability of 15 Hz global positioning system units for assessing the activity profiles of university football players | |
| Gómez et al. | Multivariate analysis of ball possessions effectiveness in elite futsal | |
| CN110348615B (en) | Cable line fault probability prediction method based on ant colony optimization support vector machine | |
| CN110458482A (en) | A kind of Evaluation of Fire Protection ability construction method based on big data | |
| CN108495329B (en) | Method and device for evaluating reliability of base station | |
| CN109586952A (en) | Method of server expansion, device | |
| Mizuguchi et al. | A new approach to determining net impulse and identification of its characteristics in countermovement jumping: Reliability and validity | |
| CN107335220B (en) | Negative user identification method and device and server | |
| CN110162445A (en) | The host health assessment method and device of Intrusion Detection based on host log and performance indicator | |
| Chen et al. | The dynamics of performance space of Major League Baseball pitchers 1871–2006 | |
| CN111754241A (en) | User behavior perception method, device, equipment and medium | |
| CN107547266A (en) | The detection method and device of online amount abnormity point, computer equipment and storage medium | |
| Corbett et al. | Weak relationships between stint duration, physical and skilled match performance in Australian football | |
| CN110795324A (en) | Data processing method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into 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: 20160420 |