CN106845722A - A kind of method and apparatus for predicting customer volume - Google Patents
A kind of method and apparatus for predicting customer volume Download PDFInfo
- Publication number
- CN106845722A CN106845722A CN201710065604.8A CN201710065604A CN106845722A CN 106845722 A CN106845722 A CN 106845722A CN 201710065604 A CN201710065604 A CN 201710065604A CN 106845722 A CN106845722 A CN 106845722A
- Authority
- CN
- China
- Prior art keywords
- application program
- target period
- cycle
- prediction
- user
- 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
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"
Abstract
The invention discloses a kind of method and apparatus for predicting customer volume, belong to field of computer technology.Methods described includes：Actual user according to application program in the reference cycle of target period measures the prediction churn rate with the target period relative to the reference cycle, predict user decrement of the target period relative to the reference cycle, wherein, the reference cycle was the upper cycle adjacent with the target period, according to application program the target period prediction installation and application program the target period the effective installation rate of prediction, predict user increment of the target period relative to the reference cycle, according to user's decrement and user's increment, and actual user's amount in the reference cycle, customer volume of the prediction application program in the target period.Using the present invention, customer volume can be relatively accurately predicted.
Description
Technical field
The present invention relates to field of computer technology, more particularly to a kind of method and apparatus for predicting customer volume.
Background technology
With continuing to develop for computer technology, increasing application program is developed and used.In application journey
After sequence listing, the exploitation side of application program is frequently necessary to estimate customer volume, before the market recent to understand application program
Scape, so as to adjust the Promotion Strategy of application program.
Exploitation side can records application program historic user amount, and by historic user amount it is corresponding with historical date storage.
Fortune can be fitted by the way of returning using historical date and historic user amount as dependent variable and independent variable afterwards
Calculate, determine the relational model of date and customer volume.And then, exploitation side can be based on the relational model, with reference to the history of record
Customer volume and historical date, complete estimating to the future customer amount of future date.
Realize it is of the invention during, inventor find prior art at least there is problems with：
The future customer amount estimated out by the way of regression fit is only capable of reflecting the trend of future customer amount ups and downs, and nothing
The reason for method explains future customer amount ups and downs exactly, and then the exploitation side of application program effectively cannot also enter to Promotion Strategy
Row specific aim is adjusted, and so, the effect of customer volume prediction is relatively low.
The content of the invention
In order to solve problem of the prior art, a kind of method and apparatus for predicting customer volume are the embodiment of the invention provides.
The technical scheme is as follows：
First aspect, there is provided a kind of method of prediction customer volume, methods described includes：
According to application program in actual user's amount in the reference cycle of target period and the target period relative to described
The prediction churn rate in reference cycle, predicts user decrement of the target period relative to the reference cycle, wherein, institute
It was the upper cycle adjacent with the target period to state the reference cycle；
According to application program the target period prediction installation and application program the target period prediction
Effective installation rate, predicts user increment of the target period relative to the reference cycle；
Actual user according to user's decrement and user's increment, and the reference cycle measures, prediction application
Customer volume of the program in the target period.
Second aspect, there is provided a kind of device of prediction customer volume, described device includes：
First prediction module, measures and the mesh for the actual user according to application program in the reference cycle of target period
Mark cycle phase predicts the target period relative to the reference cycle for the prediction churn rate in the reference cycle
User's decrement, wherein, the reference cycle was the upper cycle adjacent with the target period；
Second prediction module, for according to application program the target period prediction installation and application program in institute
The effective installation rate of prediction of target period is stated, user increment of the target period relative to the reference cycle is predicted；
3rd prediction module, for according to user's decrement and user's increment, and the reference cycle reality
Border customer volume, customer volume of the prediction application program in the target period.
The beneficial effect that technical scheme provided in an embodiment of the present invention is brought is：
In the embodiment of the present invention, when the customer volume of application program is predicted, it is considered to churn rate, the installation of application program
The factor of the various influence customer volumes of amount and effectively installation rate etc., can relatively accurately be predicted customer volume, while can
The reason for reasonably to explain customer volume ups and downs according to one of the above or many factors.
Brief description of the drawings
Technical scheme in order to illustrate more clearly the embodiments of the present invention, below will be to that will make needed for embodiment description
Accompanying drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for
For those of ordinary skill in the art, on the premise of not paying creative work, other can also be obtained according to these accompanying drawings
Accompanying drawing.
Fig. 1 is a kind of method flow diagram for predicting customer volume provided in an embodiment of the present invention；
Fig. 2 is a kind of change in long term schematic diagram of churn rate provided in an embodiment of the present invention；
Fig. 3 is a kind of principle schematic for predicting customer volume provided in an embodiment of the present invention；
Fig. 4 is a kind of method flow for determining expected installation and the effective installation rate of expection provided in an embodiment of the present invention
Figure；
Fig. 5 is a kind of method flow diagram for determining prospective users turnover rate provided in an embodiment of the present invention；
Fig. 6 is a kind of apparatus structure schematic diagram for predicting customer volume provided in an embodiment of the present invention；
Fig. 7 is a kind of structural representation of network equipment provided in an embodiment of the present invention.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with accompanying drawing to embodiment party of the present invention
Formula is described in further detail.
A kind of method for predicting customer volume is the embodiment of the invention provides, this method is mainly used in application issued
Afterwards, the exploitation root of application program according to application program in historical period customer volume, to the application program in following certain time
Customer volume be predicted so that based on the application program for predicting customer volume adjustment application program market strategy scene
Under.The method can be realized that the network equipment can be the network equipment for predicting the customer volume of application program by the network equipment,
Can be the background server of above-mentioned application program, or exploitation side is specifically used to the equipment of predicting customer volume.Network sets
Standby to include processor, memory, processor can be used for carrying out the treatment of the prediction customer volume in following flows, memory
Can be used for storing the data of the data and generation needed in following processing procedures.In the present embodiment, with the network equipment as should
Illustrated with as a example by the background server of program, other situations are similar to therewith, and this implementation is no longer illustrated one by one.
Below in conjunction with specific embodiment, the handling process shown in Fig. 1 is described in detail, content can be as
Under：
Step 101, actual user amount and target week of the network equipment according to application program in the reference cycle of target period
Phase, relative to the prediction churn rate in reference cycle, predicts user decrement of the target period relative to the reference cycle.
Wherein, target period is the future time section of customer volume to be predicted；Customer volume can be opened in certain period of time
The total amount of the account of dynamic application program；Target period can deposit in the reference cycle relative to the churn rate in reference cycle
, and the quantity of non-existent user accounts for the ratio that the actual user in reference cycle measures in target period；Target period relative to
User's decrement in reference cycle can be existed in the reference cycle, and in target period non-existent user quantity；With reference to week
Phase was the upper cycle adjacent with target period.
In force, after product (i.e. application program) listing, product development side is generally needed according to the recent reality of product
Border customer volume estimates the future customer amount tendency of product, the city current such that it is able to adjust product according to the result estimated
Field strategy.First, the change of product user magnitude may be considered in the short term in the presence of periodically, regular, such as week
End and working day can alternately change, and annual festivals or holidays can all have similar different from change on ordinary days etc., so in prediction
During customer volume, cyclic forecast can be carried out, that is, divide time into the cycle of equal length, a cycle can be a star
Phase, one month etc..Secondly, the future trend of product user amount is only dependent upon " the current customer volume of product ", and " future leaves this
The user of product ", " future newly adds the user of the product ", so can be measured by the actual user in a upper cycle, next cycle
The customer volume in next cycle was predicted relative to three amounts of user's decrement and user's increment in a upper cycle.Product development side is permissible
The prediction work of future customer amount is completed by the network equipment, specifically, the network equipment can be obtained and the log history period
Actual user's amount, then can be measured with reference to the actual user in the cycle in the finish time in certain cycle (i.e. reference cycle)
To predict the customer volume of next cycle (i.e. target period), for example, a week can be one week, Monday is the cycle,
Sunday is end cycle, then product development side can predict next week with reference to the customer volume in this cycle on Sunday
The customer volume of phase.
When user starts simultaneously login application program every time in reference cycle, (i.e. network sets the background server of application program
It is standby) can once be recorded, so, the network equipment can record the actual startup amount and reality of application program in the reference cycle
Customer volume.During prediction, actual user's amount that the network equipment can obtain the reference cycle of record (is tied to the reference cycle
At the beam moment, the total amount of the user of above-mentioned application program was started in the reference cycle), it will be understood that same user is in the reference cycle
Interior startup is applied multiple times program, and its corresponding customer volume is 1, if a certain user is only mounted with to apply journey within the reference cycle
Sequence, and not actuated application program, its corresponding customer volume are 0.May then based on target period pre- relative to the reference cycle
Survey churn rate, the quantity that the user come in the prediction reference cycle can be lost in target period, i.e. target period relative to
User's decrement in reference cycle.
Optionally, the churn rate of target period can be predicted with time value according to history, corresponding treatment can be as
Under：The network equipment determines the history cycle for having identical date feature with target period, and history cycle is gone through relative to corresponding
The churn rate in history reference cycle, is defined as prediction churn rate of the target period relative to the reference cycle.
Wherein, identical date feature can be belonging to same period in red-letter day or the working day and holiday by equal number
Composition, for example, target period is vacation 7 day National Day in 2017, then history cycle can be vacation 7 day National Day in 2015；And example
Such as, target period is the first month of the lunar year junior one of Chinese zodiac rooster year to the seventh day of lunar month, then history cycle can be the first month of the lunar year junior one of Chinese zodiac monkey year at the beginning of
Seven；Again for example, target period is the 3rd week of April, then history cycle can be 4 second weeks of month.
In force, the network equipment predict target period with respect to the reference cycle user's decrement when, it is necessary to first determine mesh
Prediction churn rate of the mark cycle phase for the reference cycle.And assume that user in the week for possessing identical date feature first
Behavior in phase is similar to, use habit is identical, so because user's custom is basically identical, it may be determined that it is special possessing phase same date
The wastage of user is consistent between the cycle levied, then can first determine the history week for having identical date feature with target period
Phase, history cycle is then defined as target period relative to ginseng relative to the churn rate in corresponding history reference cycle
Examine the prediction churn rate in cycle.It should be noted that for churn rate, in specific business diagnosis, this data
It is in fact highly stable, as shown in Fig. 2 only 1~2% fluctuation on long terms, so meet by taking history with time value
Forecast demand.
Step 102, the network equipment is all in target in the prediction installation and application program of target period according to application program
The effective installation rate of prediction of phase, user increment of the prediction target period relative to the reference cycle.
Wherein, effective installation rate can be the ratio of user's increment and installation in a certain cycle；Target period relative to
User's increment in reference cycle can not existed in the reference cycle, and the quantity of the user increased newly in target period.
In force, during customer volume is predicted, the network equipment can be predicted relative to reference cycle, target period
The quantity of interior newly-increased user, this certain customers can include newly being installed and activated in target period the use of the application program
Family, it is also possible to including being mounted with the application program, but it is not actuated within the reference cycle, and starting in target period to answer
With the user of program.Specifically, installation that can be first to application program in target period is predicted, then again to target week
Effective installation rate of application program is predicted in phase, obtains the prediction installation and the effective installation rate of prediction of target period, it
User increment of the target period relative to the reference cycle can be predicted according to above-mentioned two predicted value afterwards, installation will be predicted
Be multiplied the user's increment predicted with the effective installation rate of prediction.
Optionally, the installation of application program can be predicted according to product promotion scale amount, corresponding treatment can be as
Under：The network equipment according to application program target period popularization scale amount and application program the reference cycle popularization scale
Amount, and application program is in the actual installation amount in reference cycle, determines prediction installation of the application program in target period.
Wherein, the quantized value that scale amount can be the popularization scale for reactive applications program is promoted, can be specifically embodied
Quantity, the frequency of popularization, scope of popularization for partner etc..
In force, the network equipment can first determine mesh when user's increment in target period relative reference cycle is predicted
Prediction installation of the mark cycle phase for the reference cycle.It is appreciated that the popularization power of the installation of application program and application program
Degree, scale are relevant, that is, promote that scale is bigger, and installation is higher, so, it is determined that target period prediction installation when, can be with
The popularization scale amount in first comparison object cycle and the popularization scale amount in reference cycle, may then based on comparative result, according to ginseng
The installation for examining the cycle determines the prediction installation of target period.If specifically, the popularization scale amount and ginseng of target period
The popularization scale amount for examining the cycle is identical, then directly can install the prediction that the installation in reference cycle is defined as target period
Amount, and if the popularization scale amount of target period compares the popularization scale amount many 10% in reference cycle, then can be with target period
Prediction installation is 1.1 times of the installation in reference cycle.
Optionally, effective installation rate of target period, phase can be predicted with time value and product promotion mode according to history
The treatment answered can be as follows：The network equipment determines the history cycle for having identical date feature with target period, and journey is applied in acquisition
Sequence target period the way of promotion and application program history cycle the way of promotion, if application program is in target period
The way of promotion is identical in the way of promotion of history cycle with application program, then effective installation rate of history cycle is defined as into target
The effective installation rate of prediction in cycle, if application program is in the way of promotion and application program the pushing away in history cycle of target period
Wide mode is different, then the effective installation rate by application program in the reference cycle is defined as application program has in the prediction of target period
Effect installation rate.
In force, the network equipment predict target period with respect to the reference cycle user's increment when, it is necessary to first determine mesh
Mark cycle phase is for the effective installation rate of the prediction in reference cycle.It is possible, firstly, to first determine have phase same date special with target period
The history cycle levied, then obtains the way of promotion of target period and the way of promotion of history cycle, if target period is pushed away
Wide mode is identical with the way of promotion of history cycle, then effective installation rate of history cycle can be defined as into the pre- of target period
Effective installation rate is surveyed, if the way of promotion of target period is different with the way of promotion of history cycle, can be by the reference cycle
Effective installation rate be defined as the effective installation rate of prediction of target period.Further, the network equipment can with log history when
The corresponding effective installation rate of the difference way of promotion, then generates the corresponding relation of the way of promotion and effective installation rate in section, so,
It is determined that during the effective installation rate of the prediction of target period, the popularization side of target period can be directly determined according to above-mentioned corresponding relation
The corresponding effective installation rate of formula.
Step 103, the network equipment is measured according to user's decrement and user's increment, and the actual user in reference cycle, prediction
Customer volume of the application program in target period.
In force, after prediction completes user increment and user decrement of the target period relative to the reference cycle, can be with
Actual user according to user's decrement and user's increment, and reference cycle measures, and predicts the customer volume of target period, i.e. target week
User increment-target period phase of the actual user's amount+target period in the customer volume=reference cycle of phase relative to the reference cycle
For user's decrement in reference cycle.As shown in figure 3, wherein, when the cycle two being target period, then the cycle one is with reference to week
Phase, the cycle one user pond be actual user's amount in reference cycle, the retention amount in cycle one is then the actual user in reference cycle
Amount subtracts user decrement of the target period relative to the reference cycle, and the influx of cycle two is then target period relative to the reference cycle
User's increment, the cycle two user pond then for target period customer volume；And the cycle two is the reference cycle, the cycle three is target week
Phase, related notion may be referred to described above.
In the embodiment of the present invention, for the target period of customer volume to be predicted, the reference cycle was its adjacent upper cycle,
During customer volume is predicted, actual user's amount and target week that the network equipment can first according to application program in the reference cycle
Phase, relative to the prediction churn rate in reference cycle, predicts user decrement of the target period relative to the reference cycle, Zhi Houke
With further according to application program target period prediction installation and application program target period the effective installation rate of prediction, in advance
User increment of the target period relative to the reference cycle is surveyed, finally can be according to user's decrement and user's increment, and reference week
Actual user's amount of phase, customer volume of the prediction application program in target period.So, when the customer volume of application program is predicted,
Consider the factor of various influence customer volumes such as churn rate, the installation of application program and effective installation rate, can to
Family amount is relatively accurately predicted, while the original of customer volume ups and downs can reasonably be explained according to one of the above or many factors
Cause.
Based on identical technology design, the present embodiment also discloses the pre- measuring and calculating that a kind of network equipment is based on above-mentioned customer volume
Method, the prospective users amount of the target period set according to application development side, determines expection of the application program in target period
Installation and/or the effective installation rate of expection in target period, corresponding treatment can flows as shown in Figure 4：
Step 401, actual user amount and target week of the network equipment according to application program in the reference cycle of target period
Phase, relative to the prediction churn rate in reference cycle, predicts user decrement of the target period relative to the reference cycle.
In force, when user starts simultaneously login application program every time in the reference cycle, the network equipment can be carried out once
Record, so, the network equipment can record the actual startup amount of application program and actual user in the reference cycle and measure.Predicting
Cheng Zhong, the network equipment can obtain actual user's amount in the reference cycle of record, may then based on target period relative to ginseng
Examine the prediction churn rate in cycle, the quantity that the user come in the prediction reference cycle can be lost in target period, i.e. target
User decrement of the cycle phase for the reference cycle.
Step 402, the network equipment obtains the prospective users amount of the application program of target period.
In force, application development side can first set customer volume target of the application program in target period
(i.e. the prospective users amount of the application program of target period), then can by the prospective users amount typing network equipment so that
The network equipment can obtain the prospective users amount of the application program of target period.It should be noted that prospective users amount herein
Different from prediction customer volume above, prediction customer volume is application development root actual within the reference cycle according to application program
Customer volume, analysis is assumed based on rational, and the customer volume to application program in target period derive the value being calculated,
And prospective users amount is desired application development side, application program needs the reason of the customer volume for reaching in target period
Think value.
Step 403, the network equipment is all in reference according to the prospective users amount and application program of the application program of target period
Actual user's amount of phase, and the target period predicted determines the application of target period relative to user's decrement in reference cycle
The prospective users increment of program.
In force, the network equipment, can be according to this after the prospective users amount for getting the setting of application program opening side
Prospective users amount and application program the actual user in the reference cycle measure, and the target period predicted in step 401 relative to
User's decrement in reference cycle, determines the prospective users increment of the application program of target period, the i.e. prospective users of target period
User of the actual user's amount+target period in the prospective users amount-reference cycle of increment=target period relative to the reference cycle
Decrement.It is appreciated that prospective users increment herein is different from user's increment of prediction above.
Step 404, the network equipment determines answering for target period according to the prospective users increment of the application program of target period
With the effective installation rate of expection of the application program of the expected installation and target period of program.
In force, the network equipment can be based on after the prospective users increment that the application program of target period is determined
The desired value, determines expected installation of the application program in target period and the effective installation rate of expection.Based on description above
Understand, the user's increment=effective installation rate of installation *, and then be not difficult to obtain, it is contemplated that user's increment=expection installation * is contemplated that
Effect installation rate, further, the step can specifically be divided into three kinds of situations：
Situation one, the network equipment is according to the prospective users increment of the application program of target period and the application journey of target period
The prediction installation of sequence, determines the effective installation rate of expection of the application program of target period.
In force, the network equipment is determining application program after the prospective users increment of target period, can be with root
According to application program target period popularization scale amount in the reference cycle of popularization scale amount and application program, and application program
In the actual installation amount in reference cycle, prediction installation of the application program in target period is determined.So, the network equipment can be with root
According to the prospective users increment and the prediction installation of target period of target period, the effective installation rate of expection of target period is determined,
That is the prediction installation of the prospective users increment ÷ target periods of the effective installation rate=target period of expection of target period.Enter one
Step, the effective installation rate of expection of the application program of the target period that application development side can determine the network equipment,
Contrasted relative to the effective installation rate of the prediction in reference cycle with target period, if it is expected that effectively installation rate has more than prediction
Effect installation rate, then representing application development side needs to take measures to improve effective installation rate, for example, can be applied by adjusting
The way of promotion of program improves effective installation rate of application program, so, just can guarantee that use of the application program in target period
Family amount is up to standard, i.e., customer volume is not less than prospective users amount.
Situation two, the network equipment is according to the prospective users increment of the application program of target period and the application journey of target period
The effective installation rate of prediction of sequence, determines the expected installation of the application program of target period.
In force, the network equipment is determining application program after the prospective users increment of target period, can be with root
According to the scheme being described above, the effective installation rate based on history cycle and reference cycle determines that the prediction of target period is effectively installed
Rate.So, the network equipment can according to the effective installation rate of the prediction of prospective users increment and target period of target period, it is determined that
The prospective users increment ÷ target periods of the expected installation=target period of the expected installation of target period, i.e. target period
The effective installation rate of prediction.Further, the application of the target period that application development side can determine the network equipment
The expected installation of program, and target period is contrasted relative to the prediction installation in reference cycle, if it is expected that installation
More than prediction installation, then representing application development side needs to take measures to improve installation, for example, can be answered by adjustment
The installation of application program is improved with the popularization scale of program, so, user of the application program in target period is just can guarantee that
Amount is up to standard, i.e., customer volume is not less than prospective users amount.
Situation three, the network equipment determines answering for target period according to the prospective users increment of the application program of target period
With the effective installation rate of the expection of program and expected installation.
In force, the network equipment, can be directly according to target period after the prospective users increment for obtaining target period
Prospective users increment, the effective installation rate of the expection of expected installation and target period of target period is determined, specifically, network
Equipment can simultaneously adjust the installation and effective installation rate of target period, that is, ensure the expected installation and target of target period
The product of the effective installation rate of expection in cycle is equal with prospective users increment all the time, for example, it is contemplated that user's increment increases for prediction user
1.21 times of amount, then the installation of target period can be risen to 1.1 times of prediction installation, while having target period
Effect installation rate rises to 1.1 times of the effective installation rate of prediction.Further, situation one and situation two are may be referred to, if it is expected that
User's increment needs to take measures to improve target period relative to reference to week more than prediction user's increment, then application development side
User's increment of phase, you can to adjust the way of promotion of application program and the popularization scale of application program simultaneously, reach raising and use
The purpose of family increment.
In the embodiment of the present invention, the prospective users amount of the application program of target period for giving, the network equipment can be with
First prediction target period is then based on the use of prospective users amount and prediction relative to user's decrement of the application program in reference cycle
Family decrement determines prospective users increment, and then the expected peace of the application program of target period can be determined according to prospective users increment
The effective installation rate of expection of the application program of loading amount and/or target period.So, the exploitation side of application program can be by expection
Installation and the effective installation rate of expection, the market strategy of the application program of target period is carried out it is effective, targetedly adjust,
To ensure that application program reaches desired value in the customer volume of target period.
Based on identical technology design, the present embodiment also discloses the pre- measuring and calculating that a kind of network equipment is based on above-mentioned customer volume
Method, the prospective users amount of the target period set according to application development side, determines expection of the application program in target period
Churn rate, corresponding treatment can flow as shown in Figure 5：
Step 501, the network equipment is all in target in the prediction installation and application program of target period according to application program
The effective installation rate of prediction of phase, user increment of the prediction target period relative to the reference cycle.
In force, during customer volume is predicted, the network equipment can be predicted relative to reference cycle, target period
The quantity of interior newly-increased user, this certain customers can include newly being installed and activated in target period the use of the application program
Family, it is also possible to including being mounted with the application program, but it is not actuated within the reference cycle, and starting in target period to answer
With the user of program.Specifically, installation that can be first to application program in target period is predicted, then again to target week
Effective installation rate of application program is predicted in phase, obtains the prediction installation and the effective installation rate of prediction of target period, it
User increment of the target period relative to the reference cycle can be predicted according to above-mentioned two predicted value afterwards, installation will be predicted
Be multiplied the user's increment predicted with the effective installation rate of prediction.
Step 502, the network equipment obtains the prospective users amount of the application program of target period.
In force, application development side can first set customer volume target of the application program in target period
(i.e. the prospective users amount of the application program of target period), then can by the prospective users amount typing network equipment so that
The network equipment can obtain the prospective users amount of the application program of target period.It should be noted that prospective users amount herein
Different from prediction customer volume above, prediction customer volume is application development root actual within the reference cycle according to application program
Customer volume, analysis is assumed based on rational, and the customer volume to application program in target period derive the value being calculated,
And prospective users amount is desired application development side, application program needs the reason of the customer volume for reaching in target period
Think value.
Step 503, the network equipment is all in reference according to the prospective users amount and application program of the application program of target period
Actual user's amount of phase, and the target period predicted determines the application of target period relative to user's increment in reference cycle
The prospective users decrement of program.
In force, the network equipment, can be according to this after the prospective users amount for getting the setting of application program opening side
Prospective users amount and application program the actual user in the reference cycle measure, and the target period predicted in step 401 relative to
User's increment in reference cycle, determines the prospective users decrement of the application program of target period, the i.e. prospective users of target period
User of the actual user's amount+target period in the prospective users amount-reference cycle of decrement=target period relative to the reference cycle
Increment.It is appreciated that prospective users decrement herein is different from user's decrement of prediction above.
Step 504, the network equipment is being referred to according to the prospective users decrement and application program of the application program of target period
Actual user's amount in cycle, determines the prospective users turnover rate of the application program of target period.
In force, the network equipment, can basis after the prospective users decrement that the application program of target period is determined
The desired value and application program the actual user in the reference cycle measure, and the prospective users turnover rate of target period is determined, based on preceding
Knowable to the description of text, the reality in the prospective users decrement ÷ reference cycles of the prospective users turnover rate=target period of target period
Customer volume.Further, the expection of the application program of the target period that application development side can determine the network equipment
Churn rate, and target period is contrasted relative to the prediction churn rate in reference cycle, if it is expected that turnover rate is small
In prediction churn rate, then representing application development side needs to take measures to reduce churn rate, for example, can pass through
Upgrade in time, add the modes such as New function to reduce the loss of user, so, just can guarantee that use of the application program in target period
Family amount is up to standard, i.e., customer volume is not less than prospective users amount.
In the embodiment of the present invention, the prospective users amount of the application program of target period for giving, the network equipment can be with
First prediction target period is then based on the use of prospective users amount and prediction relative to user's increment of the application program in reference cycle
Family increment determines prospective users decrement, and then can be measured according to the actual user in prospective users decrement and reference cycle, determines mesh
The prospective users turnover rate of the application program in mark cycle.So, the exploitation side of application program can by prospective users turnover rate,
The market strategy of the application program of target period is carried out it is effective, targetedly adjust, to ensure application program in target week
The customer volume of phase reaches desired value.
Based on identical technology design, the embodiment of the present invention additionally provides a kind of device for predicting customer volume, such as Fig. 6 institutes
Show, the device includes：
First prediction module 601, measures and institute for the actual user according to application program in the reference cycle of target period
Prediction churn rate of the target period relative to the reference cycle is stated, predicts the target period relative to described with reference to week
User's decrement of phase, wherein, the reference cycle was the upper cycle adjacent with the target period；
Second prediction module 602, for according to application program the target period prediction installation and application program
In the effective installation rate of the prediction of the target period, user increment of the target period relative to the reference cycle is predicted；
3rd prediction module 603, for according to user's decrement and user's increment, and the reference cycle
Actual user measures, customer volume of the prediction application program in the target period.
Optionally, described device also includes：
First determining module, for the history cycle for determining there is identical date feature with the target period, will be described
History cycle is defined as the target period relative to the reference relative to the churn rate in corresponding history reference cycle
The prediction churn rate in cycle, wherein, the identical date feature is to belong to the same period in red-letter day or by equal number
Working day and holiday constitute.
Optionally, described device also includes：
Second determining module, for according to application program the target period popularization scale amount and application program in institute
The popularization scale amount in reference cycle is stated, and application program determines that application program exists in the actual installation amount in the reference cycle
The prediction installation of the target period.
Optionally, described device also includes：
3rd determining module, for the history cycle for determining to have identical date feature with the target period, obtaining should
With program the target period the way of promotion and application program the history cycle the way of promotion, if application program
It is identical in the way of promotion of the history cycle with application program in the way of promotion of the target period, then application program is existed
Effective installation rate of the history cycle is defined as prediction effective installation rate of the application program in the target period, if using
Program is different in the way of promotion of the history cycle with application program in the way of promotion of the target period, then will be using journey
Effective installation rate of the sequence in the reference cycle is defined as prediction effective installation rate of the application program in the target period.
Optionally, described device also includes：
4th determining module, the prospective users amount of the application program for obtaining the target period, according to the target
Actual user amount of the prospective users amount and application program of the application program in cycle in the reference cycle, and prediction is described
Target period determines the prospective users increment of the application program of target period, root relative to user's decrement in the reference cycle
According to the prospective users increment of the application program of the target period, the expected installation of the application program of the target period is determined
With the effective installation rate of the expection of the application program of the target period.
Optionally, described device also includes：
5th determining module, the prospective users amount of the application program for obtaining the target period, according to the target
Actual user amount of the prospective users amount and application program of the application program in cycle in the reference cycle, and prediction is described
Target period determines the prospective users decrement of the application program of target period, root relative to user's increment in the reference cycle
Actual user according to the prospective users decrement and application program of the application program of the target period in the reference cycle measures, really
The prospective users turnover rate of the application program of the fixed target period.
In the embodiment of the present invention, for the target period of customer volume to be predicted, the reference cycle was its adjacent upper cycle,
During customer volume is predicted, actual user's amount and target week that the network equipment can first according to application program in the reference cycle
Phase, relative to the prediction churn rate in reference cycle, predicts user decrement of the target period relative to the reference cycle, Zhi Houke
With further according to application program target period prediction installation and application program target period the effective installation rate of prediction, in advance
User increment of the target period relative to the reference cycle is surveyed, finally can be according to user's decrement and user's increment, and reference week
Actual user's amount of phase, customer volume of the prediction application program in target period.So, when the customer volume of application program is predicted,
Consider the factor of various influence customer volumes such as churn rate, the installation of application program and effective installation rate, can to
Family amount is relatively accurately predicted, while the original of customer volume ups and downs can reasonably be explained according to one of the above or many factors
Cause.
It should be noted that：Above-described embodiment provide prediction customer volume device predict customer volume when, only with above-mentioned
The division of each functional module is carried out for example, in practical application, as needed can distribute by different above-mentioned functions
Functional module is completed, will the internal structure of device be divided into different functional modules, with complete it is described above whole or
Partial function.In addition, the device of the prediction customer volume of above-described embodiment offer belongs to same with the embodiment of the method for prediction customer volume
One design, it implements process and refers to embodiment of the method, repeats no more here.
Fig. 7 is the structural representation of the network equipment provided in an embodiment of the present invention.The network equipment 700 can be because of configuration or property
Can the different and larger difference of producing ratio, one or more central processing units (central processing can be included
Units, CPU) 722 (for example, one or more processors) and memory 732, one or more storages apply journey
The storage medium 730 (such as one or more mass memory units) of sequence 742 or data 744.Wherein, the He of memory 732
Storage medium 730 can be of short duration storage or persistently storage.The program stored in storage medium 730 can include one or one
With upper module (diagram is not marked), each module can be included to the series of instructions operation in the network equipment.Further,
Central processing unit 722 could be arranged to be communicated with storage medium 730, and in performing storage medium 730 on the network equipment 700 one
Series of instructions is operated.
The network equipment 700 can also include one or more power supplys 726, one or more wired or wireless nets
Network interface 750, one or more input/output interfaces 758, one or more keyboards 756, and/or, one or one
Individual above operating system 741, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM etc.
Deng.
The network equipment 700 can include memory, and one or more than one program, one of them or one
Individual procedure above is stored in memory, and is configured to one or one by one or more than one computing device
Individual procedure above includes the instruction for carrying out above-mentioned prediction customer volume.
One of ordinary skill in the art will appreciate that realizing that all or part of step of above-described embodiment can be by hardware
To complete, it is also possible to instruct the hardware of correlation to complete by program, described program can be stored in a kind of computer-readable
In storage medium, storage medium mentioned above can be read-only storage, disk or CD etc..
The foregoing is only presently preferred embodiments of the present invention, be not intended to limit the invention, it is all it is of the invention spirit and
Within principle, any modification, equivalent substitution and improvements made etc. should be included within the scope of the present invention.
Claims (12)
1. it is a kind of predict customer volume method, it is characterised in that methods described includes：
According to application program in actual user's amount in the reference cycle of target period and the target period relative to the reference
The prediction churn rate in cycle, predicts user decrement of the target period relative to the reference cycle, wherein, the ginseng
The cycle of examining was the upper cycle adjacent with the target period；
It is effective in the prediction of the target period in the prediction installation and application program of the target period according to application program
Installation rate, predicts user increment of the target period relative to the reference cycle；
Actual user according to user's decrement and user's increment, and the reference cycle measures, and predicts application program
In the customer volume of the target period.
2. method according to claim 1, it is characterised in that methods described also includes：
It is determined that having the history cycle of identical date feature with the target period, the history cycle is gone through relative to corresponding
The churn rate in history reference cycle, is defined as prediction churn rate of the target period relative to the reference cycle,
Wherein, the identical date feature is to belong to the same period in red-letter day or be made up of the working day and holiday of equal number.
3. method according to claim 1, it is characterised in that methods described also includes：
According to application program the target period popularization scale amount and application program the reference cycle popularization scale
Amount, and application program is in the actual installation amount in the reference cycle, determines that application program is pacified in the prediction of the target period
Loading amount.
4. method according to claim 1, it is characterised in that methods described also includes：
It is determined that having the history cycle of identical date feature with the target period；
Obtain application program the target period the way of promotion and application program the history cycle the way of promotion；
If application program the target period the way of promotion and application program the history cycle way of promotion phase
Together, then application program is defined as into application program in effective installation rate of the history cycle has in the prediction of the target period
Effect installation rate；
If application program the target period the way of promotion and application program the history cycle the way of promotion not
Together, then the effective installation rate by application program in the reference cycle is defined as application program has in the prediction of the target period
Effect installation rate.
5. the method according to claim any one of 1-4, it is characterised in that methods described also includes：
The prospective users amount of the application program of the target period is obtained, application program according to the target period is expected with
Family is measured and actual user of the application program in the reference cycle measures, and the target period predicted is relative to the reference
User's decrement in cycle, determines the prospective users increment of the application program of target period；
The prospective users increment of the application program according to the target period, determines the expection of the application program of the target period
The effective installation rate of expection of the application program of installation and the target period.
6. the method according to claim any one of 1-4, it is characterised in that methods described also includes：
The prospective users amount of the application program of the target period is obtained, application program according to the target period is expected with
Family is measured and actual user of the application program in the reference cycle measures, and the target period predicted is relative to the reference
User's increment in cycle, determines the prospective users decrement of the application program of target period；
Actual use of the prospective users decrement and application program of the application program according to the target period in the reference cycle
Family is measured, and determines the prospective users turnover rate of the application program of the target period.
7. it is a kind of predict customer volume device, it is characterised in that described device includes：
First prediction module, for the actual user's amount according to application program in the reference cycle of target period and target week
Phase, relative to the prediction churn rate in the reference cycle, predicts user of the target period relative to the reference cycle
Decrement, wherein, the reference cycle was the upper cycle adjacent with the target period；
Second prediction module, for according to application program the target period prediction installation and application program in the mesh
The effective installation rate of prediction in mark cycle, predicts user increment of the target period relative to the reference cycle；
3rd prediction module, for according to user's decrement and user's increment, and the reference cycle actual use
Family is measured, customer volume of the prediction application program in the target period.
8. device according to claim 7, it is characterised in that described device also includes：
First determining module, for the history cycle for determining there is identical date feature with the target period, by the history
Cycle phase is defined as the target period relative to the reference cycle for the churn rate in corresponding history reference cycle
Prediction churn rate, wherein, the identical date feature is to belong to same period in red-letter day or the work by equal number
Day and holiday composition.
9. device according to claim 7, it is characterised in that described device also includes：
Second determining module, for according to application program the target period popularization scale amount and application program in the ginseng
The popularization scale amount in cycle is examined, and application program determines application program described in the actual installation amount in the reference cycle
The prediction installation of target period.
10. device according to claim 7, it is characterised in that described device also includes：
3rd determining module, for the history cycle for determining to have identical date feature with the target period, journey is applied in acquisition
Sequence the target period the way of promotion and application program the history cycle the way of promotion, if application program is in institute
The way of promotion for stating target period is identical in the way of promotion of the history cycle with application program, then by application program described
Effective installation rate of history cycle is defined as prediction effective installation rate of the application program in the target period, if application program
It is different in the way of promotion of the history cycle with application program in the way of promotion of the target period, then application program is existed
Effective installation rate in the reference cycle is defined as prediction effective installation rate of the application program in the target period.
11. device according to claim any one of 7-10, it is characterised in that described device also includes：
4th determining module, the prospective users amount of the application program for obtaining the target period, according to the target period
Application program prospective users amount and application program the reference cycle actual user measure, and prediction the target
Cycle phase determines the prospective users increment of the application program of target period, according to institute for user's decrement in the reference cycle
The prospective users increment of the application program of target period is stated, expected installation and the institute of the application program of the target period is determined
State the effective installation rate of expection of the application program of target period.
12. device according to claim any one of 7-10, it is characterised in that described device also includes：
5th determining module, the prospective users amount of the application program for obtaining the target period, according to the target period
Application program prospective users amount and application program the reference cycle actual user measure, and prediction the target
Cycle phase determines the prospective users decrement of the application program of target period, according to institute for user's increment in the reference cycle
State the actual user of the prospective users decrement and application program of the application program of target period in the reference cycle to measure, determine institute
State the prospective users turnover rate of the application program of target period.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710065604.8A CN106845722A (en) | 2017-02-06 | 2017-02-06 | A kind of method and apparatus for predicting customer volume |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710065604.8A CN106845722A (en) | 2017-02-06 | 2017-02-06 | A kind of method and apparatus for predicting customer volume |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106845722A true CN106845722A (en) | 2017-06-13 |
Family
ID=59122982
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710065604.8A Pending CN106845722A (en) | 2017-02-06 | 2017-02-06 | A kind of method and apparatus for predicting customer volume |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106845722A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108364195A (en) * | 2018-02-09 | 2018-08-03 | 腾讯科技（深圳）有限公司 | User retains probability forecasting method, device, predictive server and storage medium |
| CN108765010A (en) * | 2018-05-29 | 2018-11-06 | 腾讯科技（深圳）有限公司 | User retains probability forecasting method, device, predictive server and storage medium |
| CN109711897A (en) * | 2018-12-29 | 2019-05-03 | 贵州创鑫旅程网络技术有限公司 | Day any active ues quantity prediction technique and device |
| CN109740822A (en) * | 2019-01-22 | 2019-05-10 | 深圳市酷开网络科技有限公司 | A kind of prediction processing method that OTT device is actively measured, system and storage medium |
| CN110689381A (en) * | 2019-10-10 | 2020-01-14 | 中国联合网络通信集团有限公司 | Early warning method and device |
| CN111062749A (en) * | 2019-12-12 | 2020-04-24 | 北京爱奇艺科技有限公司 | Growth amount estimation method, growth amount estimation device, electronic apparatus, and storage medium |
| CN112905191A (en) * | 2019-11-19 | 2021-06-04 | 腾讯科技（深圳）有限公司 | Data processing method, data processing device, computer readable storage medium and computer equipment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070185867A1 (en) * | 2006-02-03 | 2007-08-09 | Matteo Maga | Statistical modeling methods for determining customer distribution by churn probability within a customer population |
| CN102395135A (en) * | 2011-10-25 | 2012-03-28 | 江苏省邮电规划设计院有限责任公司 | VLR (Visitor Location Register) user number predicting method based on gray system model neural network |
| CN104504084A (en) * | 2014-12-24 | 2015-04-08 | 小米科技有限责任公司 | Method and device for determining user retention rate |
| CN106022511A (en) * | 2016-05-11 | 2016-10-12 | 北京京东尚科信息技术有限公司 | Information predicting method and device |
| CN106056400A (en) * | 2016-05-19 | 2016-10-26 | 政和科技股份有限公司 | Method, apparatus and system of predicting number of new users |
| CN106251017A (en) * | 2016-08-02 | 2016-12-21 | 北京百度网讯科技有限公司 | Data predication method and device |
| CN106250999A (en) * | 2015-06-03 | 2016-12-21 | 阿里巴巴集团控股有限公司 | The methods, devices and systems of prediction turnover rate |
-
2017
- 2017-02-06 CN CN201710065604.8A patent/CN106845722A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070185867A1 (en) * | 2006-02-03 | 2007-08-09 | Matteo Maga | Statistical modeling methods for determining customer distribution by churn probability within a customer population |
| CN102395135A (en) * | 2011-10-25 | 2012-03-28 | 江苏省邮电规划设计院有限责任公司 | VLR (Visitor Location Register) user number predicting method based on gray system model neural network |
| CN104504084A (en) * | 2014-12-24 | 2015-04-08 | 小米科技有限责任公司 | Method and device for determining user retention rate |
| CN106250999A (en) * | 2015-06-03 | 2016-12-21 | 阿里巴巴集团控股有限公司 | The methods, devices and systems of prediction turnover rate |
| CN106022511A (en) * | 2016-05-11 | 2016-10-12 | 北京京东尚科信息技术有限公司 | Information predicting method and device |
| CN106056400A (en) * | 2016-05-19 | 2016-10-26 | 政和科技股份有限公司 | Method, apparatus and system of predicting number of new users |
| CN106251017A (en) * | 2016-08-02 | 2016-12-21 | 北京百度网讯科技有限公司 | Data predication method and device |
Non-Patent Citations (1)
| Title |
|---|
| 李双杰、谭振龙、李薇、汤燕娟: "基于用户流失率的4G用户精细化预测模型研究", 《互联网天地》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108364195A (en) * | 2018-02-09 | 2018-08-03 | 腾讯科技（深圳）有限公司 | User retains probability forecasting method, device, predictive server and storage medium |
| CN108364195B (en) * | 2018-02-09 | 2021-11-23 | 腾讯科技（深圳）有限公司 | User retention probability prediction method and device, prediction server and storage medium |
| CN108765010A (en) * | 2018-05-29 | 2018-11-06 | 腾讯科技（深圳）有限公司 | User retains probability forecasting method, device, predictive server and storage medium |
| CN109711897A (en) * | 2018-12-29 | 2019-05-03 | 贵州创鑫旅程网络技术有限公司 | Day any active ues quantity prediction technique and device |
| CN109740822A (en) * | 2019-01-22 | 2019-05-10 | 深圳市酷开网络科技有限公司 | A kind of prediction processing method that OTT device is actively measured, system and storage medium |
| CN110689381A (en) * | 2019-10-10 | 2020-01-14 | 中国联合网络通信集团有限公司 | Early warning method and device |
| CN112905191A (en) * | 2019-11-19 | 2021-06-04 | 腾讯科技（深圳）有限公司 | Data processing method, data processing device, computer readable storage medium and computer equipment |
| CN112905191B (en) * | 2019-11-19 | 2023-11-07 | 腾讯科技（深圳）有限公司 | Data processing method, device, computer readable storage medium and computer equipment |
| CN111062749A (en) * | 2019-12-12 | 2020-04-24 | 北京爱奇艺科技有限公司 | Growth amount estimation method, growth amount estimation device, electronic apparatus, and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106845722A (en) | A kind of method and apparatus for predicting customer volume | |
| KR102451473B1 (en) | Dynamic optimization of simulation resources | |
| Batselier et al. | Empirical evaluation of earned value management forecasting accuracy for time and cost | |
| US8732487B2 (en) | Predictive computing device power management | |
| JP5313990B2 (en) | Estimating service resource consumption based on response time | |
| US20190325370A1 (en) | Automatic demand-driven resource scaling for relational database-as-a-service | |
| WO2017045472A1 (en) | Resource prediction method and system, and capacity management apparatus | |
| US11928620B2 (en) | Method for estimating amount of task objects required to reach target completed tasks | |
| Das et al. | Performance optimization for edge-cloud serverless platforms via dynamic task placement | |
| US7421460B2 (en) | Method for determining execution of backup on a database | |
| US10545181B2 (en) | Systems and methods for apportioning power consumption | |
| Oliver et al. | An empirical approach to smartphone energy level prediction | |
| CN107515663A (en) | The method and apparatus for adjusting central processor core running frequency | |
| US20130268310A1 (en) | Scheduling system, method and computer-readable medium | |
| US20120221373A1 (en) | Estimating Business Service Responsiveness | |
| CN108459905B (en) | Resource pool capacity planning method and server | |
| Borghesi et al. | Pricing schemes for energy-efficient HPC systems: Design and exploration | |
| CN109976975A (en) | A kind of disk size prediction technique, device, electronic equipment and storage medium | |
| US10713578B2 (en) | Estimating utilization of network resources using time series data | |
| Ma et al. | Decomposition of temporal changes in construction labour productivity | |
| CN109213965B (en) | System capacity prediction method, computer readable storage medium and terminal device | |
| CN109388550B (en) | Cache hit rate determination method, device, equipment and readable storage medium | |
| US20120278038A1 (en) | Estimating monthly heating oil consumption from fiscal year oil consumption data using multiple regression and heating degree day density function | |
| TWI777154B (en) | Electronic device and method for turnover rate prediction | |
| CN115423209B (en) | Power load demand prediction method and device, electronic 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 |