CN109039691A - The method and storage medium of server, forecasting system calling amount - Google Patents
The method and storage medium of server, forecasting system calling amount Download PDFInfo
- Publication number
- CN109039691A CN109039691A CN201810554262.0A CN201810554262A CN109039691A CN 109039691 A CN109039691 A CN 109039691A CN 201810554262 A CN201810554262 A CN 201810554262A CN 109039691 A CN109039691 A CN 109039691A
- Authority
- CN
- China
- Prior art keywords
- accessed
- data
- factor
- trained
- calling amount
- 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.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/147—Network analysis or design for predicting network behaviour
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
Abstract
The present invention relates to a kind of servers, the method and storage medium of forecasting system calling amount, this method comprises: obtaining the system parameter of system to be accessed and each access system respectively, the system current status of two systems is calculated based on system parameter;If there is system current status to be greater than default similarity threshold, to be trained factor data of the factor data of the access system as the system to be accessed is obtained;If system current status is respectively less than or is equal to default similarity threshold, obtains the system of system to be accessed within a preset time and call data, call data to pre-process system, obtain stand-by system and call data;The factor data to be trained of data-acquisition system characterization factor and element factor as the system to be accessed is called based on stand-by system;Preset model is trained based on factor data to be trained, obtains the first model of prediction calling amount.The calling amount that the present invention can treat access system future is accurately predicted, to cope with system exception problem.
Description
Technical field
The present invention relates to the methods and storage of field of communication technology more particularly to a kind of server, forecasting system calling amount
Medium.
Background technique
Currently, there is a large amount of important system under large-scale internet financial company, these systems often occur by
It increases sharply in system calling amount, the problem of leading to system crash.The method for generalling use monitoring system calling amount in the prior art is
System monitoring mode, but this system monitoring mode has the following deficiencies: it is poor using system monitoring mode timeliness, going out
Existing problem uses remedial measure to have little time again, and system calling amount lacks the management of system, be unfavorable for data accumulation and
Long-term analysis.
Summary of the invention
The purpose of the present invention is to provide a kind of servers, the method and storage medium of forecasting system calling amount, it is intended to right
The calling amount of system to be accessed is accurately predicted.
To achieve the above object, the present invention provides a kind of server, the server include memory and with the storage
The processor of device connection, is stored with the processing system that can be run on the processor, the processing system in the memory
Following steps are realized when being executed by the processor:
The system parameter for obtaining system to be accessed and each access system respectively is calculated to be accessed based on the system parameter
The system current status of system and each access system;
If the system current status for needing access system and access system is greater than default similarity threshold, obtains this and connect
The factor data for entering system, using the factor data as the factor data to be trained of the system to be accessed, the factor data packet
Include time, network type, host number, registration number of users, system calling amount, type of service, festivals or holidays factor;
If the system current status of system to be accessed and each access system is respectively less than or is equal to default similarity threshold, obtain
The system of system to be accessed within a preset time is taken to call data, the system for treating access system calls data to be pre-processed,
It obtains stand-by system and calls data;
The system features factor and element factor of the data acquisition system to be accessed are called based on the stand-by system, with
The factor data to be trained of the system features factor and element factor as the system to be accessed;
Factor data to be trained based on system to be accessed is trained preset model, for predicting after being trained
First model of the calling amount of the system to be accessed.
Preferably, the system parameter includes type of service, system architecture type, deployed environment type, portfolio and use
Family amount, described the step of system to be accessed and each system current status of access system are calculated based on the system parameter, specifically
Include:
Based on the system parameter obtain respectively system to be accessed and each similar value a1 of the type of service of access system,
The similar value a2 of system architecture, the similar value a3 of deployed environment, the similar value a4 of portfolio and the similar value a5 of user volume, and obtain
Take the preset weight W1 of type of service, the preset weight W2 of system architecture type, the preset weight W3 of deployed environment type, business
Measure preset weight W4 and user volume preset weight W5, computing system similarity R=a1*W1+a2*W2+a3*W3+a4*W4+
a5*W5。
Preferably, it includes system calling amount that the system, which calls data, and the system for treating access system calls data
Pretreated step is carried out, is specifically included:
System calling amount of the system to be accessed in the first preset time is obtained, calculates the system calling amount at this
Standard variance in first preset time;
If the standard variance is 0, the system calling amount in first preset time is filtered, and obtain the system to be accessed
System calling amount of the system in the second preset time, to calculate standard side of the system calling amount in second preset time
Difference, until the standard variance is not 0, first preset time and second preset time are time in the past, described the
One preset time is not equal to second preset time;
If the standard variance is not 0, filtration system, which is called in data, to be empty data and not to meet predetermined format
The system obtained after filtering is called data to call data as stand-by system by data.
Preferably, when the processing system is executed by the processor, following steps are also realized:
Factor data to be trained successively is rejected according to preset rejecting rule, is being rejected after training factor data every time,
The model is trained using factor data to be trained remaining after rejecting, for predicting the system to be accessed after being trained
Each second model of the calling amount of system;
The corresponding each accuracy rate of calling amount for obtaining each second model prediction system to be accessed respectively, according to described
Each accuracy rate evaluates the factor data to be trained rejected.
To achieve the above object, the present invention also provides a kind of method of forecasting system calling amount, the forecasting system is called
The method of amount includes:
S1 obtains the system parameter of system to be accessed and each access system respectively, based on the system parameter calculate to
The system current status of access system and each access system;
S2, if the system current status for needing access system and access system is greater than default similarity threshold, obtaining should
The factor data of access system, it is described because of subnumber using the factor data as the factor data to be trained of the system to be accessed
According to including time, network type, host number, registration number of users, system calling amount, type of service, festivals or holidays factor;
S3, if the system current status of system to be accessed and each access system is respectively less than or is equal to default similarity threshold,
It then obtains the system of system to be accessed within a preset time and calls data, the system for treating access system calls data to be located in advance
Reason obtains stand-by system and calls data;
S4, based on the stand-by system call the data acquisition system to be accessed the system features factor and it is basic because
Son, using the system features factor and element factor as the factor data to be trained of the system to be accessed;
S5, the factor data to be trained based on system to be accessed are trained preset model, are used for after being trained
Predict the first model of the calling amount of the system to be accessed.
Preferably, the system parameter includes type of service, system architecture type, deployed environment type, portfolio and use
Family amount, the step S1 are specifically included:
Based on the system parameter obtain respectively system to be accessed and each similar value a1 of the type of service of access system,
The similar value a2 of system architecture, the similar value a3 of deployed environment, the similar value a4 of portfolio and the similar value a5 of user volume, and obtain
Take the preset weight W1 of type of service, the preset weight W2 of system architecture type, the preset weight W3 of deployed environment type, business
Measure preset weight W4 and user volume preset weight W5, computing system similarity R=a1*W1+a2*W2+a3*W3+a4*W4+
a5*W5。
Preferably, it includes system calling amount that the system, which calls data, and the step S3 is specifically included:
System calling amount of the system to be accessed in the first preset time is obtained, calculates the system calling amount at this
Standard variance in first preset time;
If the standard variance is 0, the system calling amount in first preset time is filtered, and obtain the system to be accessed
System calling amount of the system in the second preset time, to calculate standard side of the system calling amount in second preset time
Difference, until the standard variance is not 0, first preset time and second preset time are time in the past, described the
One preset time is not equal to second preset time;
If the standard variance is not 0, filtration system, which is called in data, to be empty data and not to meet predetermined format
The system obtained after filtering is called data to call data as stand-by system by data.
Preferably, after the step S5, further includes:
S6 successively rejects factor data to be trained according to preset rejecting rule, is rejecting factor data to be trained every time
Afterwards, the model is trained using factor data to be trained remaining after rejecting, for predicting that this is waiting after being trained
Enter each second model of the calling amount of system;
S7 obtains the corresponding each accuracy rate of calling amount of each second model prediction system to be accessed respectively, according to
Each accuracy rate evaluates the factor data to be trained rejected.
Preferably, after the step S7, further includes:
If having the corresponding accuracy rate of the second model to be less than presets preset numerical value, will be rejected in training second model
Factor data to be trained recommended as important factor data;
Establish the system to be accessed and the incidence relation of corresponding important factor data and preservation.
The present invention also provides a kind of computer readable storage medium, processing is stored on the computer readable storage medium
The step of system, the processing system realizes the method for above-mentioned forecasting system calling amount when being executed by processor.
The beneficial effects of the present invention are: the present invention calculates to be accessed first before the calling amount to system is predicted
The system current status of system and each access system, if system to be accessed and each access system are not similar system,
The legacy system for then treating access system calls data to carry out the pretreatment operation such as screening, and obtains stand-by system and calls data,
Then call the system features factor for obtaining the system to be accessed in data and element factor waiting as this from stand-by system
The factor data to be trained for entering system, using factor data training pattern to be trained to construct to obtain for predicting the system to be accessed
The model for following calling amount of uniting constructs prediction model using the method for big data analysis, to be accessed in this way
The calling amount in system future is accurately predicted, to cope with system exception problem;Unification is carried out to the tune usage data of system
Management is conducive to the accumulation and long-term analysis of data.
Detailed description of the invention
Fig. 1 is the optional application environment schematic diagram of each embodiment one of the invention;
Fig. 2 is the flow diagram of the method first embodiment of forecasting system calling amount of the present invention;
Fig. 3 is the flow diagram of the method second embodiment of forecasting system calling amount of the present invention.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that described herein, specific examples are only used to explain the present invention, not
For limiting the present invention.Based on the embodiments of the present invention, those of ordinary skill in the art are not before making creative work
Every other embodiment obtained is put, shall fall within the protection scope of the present invention.
It should be noted that the description for being related to " first ", " second " etc. in the present invention is used for description purposes only, and cannot
It is interpreted as its relative importance of indication or suggestion or implicitly indicates the quantity of indicated technical characteristic.Define as a result, " the
One ", the feature of " second " can explicitly or implicitly include at least one of the features.In addition, the skill between each embodiment
Art scheme can be combined with each other, but must be based on can be realized by those of ordinary skill in the art, when technical solution
Will be understood that the combination of this technical solution is not present in conjunction with there is conflicting or cannot achieve when, also not the present invention claims
Protection scope within.
As shown in fig.1, being the application environment schematic diagram of the preferred embodiment of the method for forecasting system calling amount of the present invention.
The application environment schematic diagram includes server 1 and multiple systems to be accessed 2.Server 1 can pass through network, near-field communication technology
Data interaction is carried out Deng the terminal where suitable technology and system to be accessed 2.
The server 1 be it is a kind of can according to the instruction for being previously set or store, automatic progress numerical value calculating and/or
The equipment of information processing.The server 1 can be computer, be also possible to single network server, multiple network servers
The server group of composition or the cloud being made of a large amount of hosts or network server based on cloud computing, wherein cloud computing is point
One kind that cloth calculates, a super virtual computer consisting of a loosely coupled set of computers.
In the present embodiment, server 1 may include, but be not limited only to, and depositing for connection can be in communication with each other by system bus
Reservoir 11, processor 12, network interface 13, memory 11 are stored with the processing system that can be run on the processor 12.It needs to refer to
Out, Fig. 1 illustrates only the server 1 with component 11-13, it should be understood that being not required for implementing all show
Component, the implementation that can be substituted is more or less component.
Wherein, memory 11 includes the readable storage medium storing program for executing of memory and at least one type.Inside save as the operation of server 1
Caching is provided;Readable storage medium storing program for executing can be for if flash memory, hard disk, multimedia card, card-type memory are (for example, SD or DX memory
Deng), random access storage device (RAM), static random-access memory (SRAM), read-only memory (ROM), electric erasable can compile
Journey read-only memory (EEPROM), programmable read only memory (PROM), magnetic storage, disk, CD etc. it is non-volatile
Storage medium.In some embodiments, readable storage medium storing program for executing can be the internal storage unit of server 1, such as the server 1
Hard disk;In further embodiments, which is also possible to the External memory equipment of server 1, such as
The plug-in type hard disk being equipped on server 1, intelligent memory card (Smart Media Card, SMC), secure digital (Secure
Digital, SD) card, flash card (Flash Card) etc..In the present embodiment, the readable storage medium storing program for executing of memory 11 is commonly used in
Storage is installed on the operating system and types of applications software of server 1, such as the journey of the processing system in one embodiment of the invention
Sequence code etc..In addition, memory 11 can be also used for temporarily storing the Various types of data that has exported or will export.
The processor 12 can be in some embodiments central processing unit (Central Processing Unit,
CPU), controller, microcontroller, microprocessor or other data processing chips.The processor 12 is commonly used in the control clothes
The overall operation of business device 1, such as execute control relevant to the progress data interaction of other terminals or communication and processing etc..This reality
It applies in example, the processor 12 is used to run the program code stored in the memory 11 or processing data, such as runs
Processing system etc..
The network interface 13 may include radio network interface or wired network interface, which is commonly used in
Communication connection is established between the server 1 and other electronic equipments.In the present embodiment, network interface 13 is mainly used for service
Device 1 is connected with one or more terminal devices 2, and it is logical that data transmission is established between server 1 and one or more terminal devices 2
Road and communication connection.
The processing system is stored in memory 11, is stored in including at least one computer-readable in memory 11
Instruction, at least one computer-readable instruction can be executed by processor device 12, the method to realize each embodiment of the application;With
And the function that at least one computer-readable instruction is realized according to its each section is different, can be divided into different logic moulds
Block.
In one embodiment, following steps are realized when above-mentioned processing system is executed by the processor 12:
The system parameter for obtaining system to be accessed and each access system respectively is calculated to be accessed based on the system parameter
The system current status of system and each access system;
Wherein, server provides access interface to system to be accessed, receives the system parameter and system tune of system to be accessed
With data, it is respectively used to computing system similarity and training pattern.System parameter includes type of service, system architecture type, portion
Affix one's name to environmental form, portfolio and user volume.For example, type of service includes insurance, production danger, life insurance, vehicle insurance etc.;System architecture type
Including mobile terminal framework, PC end-rack structure etc.;Deployed environment type is the region of deployment, including goes up north wide deeply etc.;Portfolio and use
Family amount is number of levels.
In one embodiment, the system current status packet of system to be accessed and each access system is calculated based on system parameter
It includes: system to be accessed and each the similar value a1 of the type of service of access system, system is obtained based on the system parameter respectively
The similar value a2 of framework, the similar value a3 of deployed environment, the similar value a4 of portfolio and the similar value a5 of user volume, and obtain industry
The preset weight W1 of service type, the preset weight W2 of system architecture type, the preset weight W3 of deployed environment type, portfolio are pre-
If weight W4 and user volume preset weight W5, computing system similarity R=a1*W1+a2*W2+a3*W3+a4*W4+a5*
W5, wherein weight represents the system parameter for the significance level of system, if weight is bigger, the system parameter is more important.
In a specific example, for type of service, if the service class of system to be accessed and some access system
Type is all identical, then the corresponding similar value of access system has been 1 to the system to be accessed with this, if type of service have 3 kinds it is identical,
For example, system to be accessed and this access system have insurance, produce danger, 3 kinds of business of life insurance, then corresponding similar value be 0.8, such as
Fruit type of service have 2 kinds it is identical, for example, system to be accessed and this access system has insurance, produces danger, then corresponding similar value
It is 0.6.For system architecture, if system to be accessed is all identical as the system architecture of some access system, this is waiting
Entering system, the corresponding similar value of access system has been 1 with this, if system architecture have it is a kind of identical, for example, system to be accessed with
The system architecture type of the access system only includes mobile terminal framework, then corresponding similar value is 0.8.For deployed environment,
System to be accessed is all identical as the deployed environment of some access system, then access system is corresponding with this for the system to be accessed
Similar value be 1, if to have 3 identical for deployed environment, for example, the region of deployment all includes the broadly area that goes up north, then corresponding phase
It is 0.8 like value, if to have 2 identical for deployed environment, for example, the region of deployment all includes Beijing and Shanghai, then it is corresponding similar
Value is 0.6.For portfolio, if system to be accessed is with some, the number of levels of the portfolio of access system is identical, for example,
It is ten million rank, then the corresponding similar value of access system has been 1 to the system to be accessed with this, if the number of levels of portfolio
Not identical, for example, system to be accessed is ten million rank, access system is million ranks, then corresponding similar value is 0.6.It is right
In user volume, if system to be accessed is with some, the number of levels of the user volume of access system is identical, the system to be accessed
The corresponding similar value of access system has been 1 with this, if the number of levels of user volume is not identical, for example, system to be accessed is thousand
Ten thousand ranks, access system is million ranks, then corresponding similar value is 0.6.
If the system current status for needing access system and access system is greater than default similarity threshold, obtains this and connect
The factor data for entering system, using the factor data as the factor data to be trained of the system to be accessed;
Wherein, default similarity threshold is, for example, 0.85.Access system is come using method identical with access system
Following calling amount is predicted, if there is the system current status of access system and the system to be accessed is greater than default similarity threshold
Factor data that access system uses can be directly used as the system to be accessed it may be considered that two systems are more similar in value
The factor data to be trained of system.
Wherein, factor data includes time, network type, host number, registration number of users, system calling amount, business
Type, festivals or holidays factor etc., which is used directly for training pattern.
If the system current status of system to be accessed and each access system is respectively less than or is equal to default similarity threshold, obtain
The system of system to be accessed within a preset time is taken to call data, the system for treating access system calls data to be pre-processed,
It obtains stand-by system and calls data;
Wherein, if in access system, default phase is respectively less than or equal to the system current status of system to be accessed
Like degree threshold value, then it is assumed that the system to be accessed and each access system are dissimilar, therefore, it is desirable to obtain in system to be accessed
Legacy system calls data, by calling data to be pre-processed to obtain stand-by system calling data system.
In one embodiment, it includes: that acquisition is described to be accessed that the system for treating access system, which calls data to carry out pretreatment,
System calling amount of the system in the first preset time, calculates standard side of the system calling amount in first preset time
Difference;If the standard variance is 0, the system calling amount in first preset time is filtered, and obtain the system to be accessed and exist
System calling amount in second preset time, to calculate standard variance of the system calling amount in second preset time;If
The standard variance is not 0, then it is the data of sky and the data for not meeting predetermined format in data that filtration system, which is called, will be filtered
The system obtained afterwards calls data to call data as stand-by system.
Wherein, it is not pretreated data that system, which calls data, comprising: system calling amount, time, network type, master
Machine quantity, registration number of users, type of service, festivals or holidays factor etc., the first preset time is past a period of time, for example,
Past one month or two months, the second preset time was different from the first preset time, for example, the first preset time is the past
Two months, then the second preset time be past one month.It is within the past period by calculating system to be accessed
The standard variance for calling amount of uniting, the stability of the system to be accessed is determined by standard variance, i.e., whether abnormal conditions occurs.
If the standard variance of system to be accessed is 0, determine that system calling amount all within this time is equal, it cannot
As the factor data in limit of consideration.For example, for 12306 seat reservation systems, if the number bought tickets daily in one month in the past
Amount is equal, then is considered as abnormal situation, system in this one month call data cannot function as in limit of consideration because of subnumber
According to may filter that system in this one month calls data.In addition the system calling amount in the second preset time is taken, is to calculate this
Standard variance of the calling amount of uniting in second preset time, and analyze whether the standard variance is 0 again, if standard variance is
0 system filtered in second preset time calls data, then the system calling amount in other times section is taken to calculate standard side
Difference, and so on, until the system calling amount that can to obtain standard variance not be 0.
If the standard variance of system calling amount of the system to be accessed in the first preset time is not 0, for this
System in one preset time calls data to be handled, comprising: it is empty data in data that filtration system, which is called, and filtering is not inconsistent
The data of predetermined format are closed, for example, the data that date format does not meet predetermined format are filtered, value type is not predetermined number
The data of Value Types are filtered, and after above-mentioned filtration treatment, are obtained stand-by system and are called data.
The system features factor and element factor of the data acquisition system to be accessed are called based on the stand-by system, with
The factor data to be trained of the system features factor and element factor as the system to be accessed;
Wherein, the system features factor and element factor that the system to be accessed is obtained in data are called from stand-by system,
Element factor is that substantially will affect the factor of the calling amount of each system to be accessed comprising time, network type, host
Quantity, registration number of users, system calling amount etc.;The system features factor is the distinctive influence system to be accessed of system to be accessed
The factor of the calling amount of system comprising type of service, festivals or holidays factor etc..The present embodiment be not limited to above system feature because
Son and element factor can select the system features factor and element factor according to the actual conditions of system to be accessed.
In a specific example, if system to be accessed be life insurance system, element factor include the time, network type,
It is host number, registration number of users, at least several in system calling amount, in the system features factor, if type of service includes
Life insurance, endowment insurance, then festivals or holidays factor is, for example, whether double 12 have whether the second of life insurance and endowment insurance kills the movable, same day Double Ninth Festival
Release life insurance and endowment insurance discount and discount number etc., these festivals or holidays factors largely influence system calling
Amount.
Factor data to be trained based on system to be accessed is trained preset model, for predicting after being trained
First model of the calling amount of the system to be accessed.
Wherein, preset model is preferably RNN neural network model.In one embodiment, above-mentioned system to be accessed is being utilized
During the factor data to be trained of system is trained RNN neural network model, it is first determined the RNN neural network model
Training number, if there is access system and the system to be accessed it is similar (for example, system current status be greater than it is a certain preset
Threshold value), then the corresponding frequency of training of model of access system can be used, if access system and the system to be accessed are all
Dissmilarity then calls the data volume of data to determine the number of training according to the stand-by system that system to be accessed provides, if to
System calls the data volume of data more, then trained number can be relatively smaller, if stand-by system calls data
Data volume is less, then trained number can be relatively more.
In one embodiment, RNN neural network model is carried out using the factor data to be trained of above-mentioned system to be accessed
Training includes: to collect remaining 20% or 25% as verifying using 75% data of factor data to be trained as training set,
After having trained RNN neural network model with training set, the RNN neural network model after training is tested using verifying collection
Card, if the accuracy rate of model is greater than preset threshold (such as 0.985), the RNN neural network model after training can be used for
Prediction can increase factor data to be trained and be trained again if accuracy rate is not more than preset threshold, until after training
The accuracy rate of RNN neural network model is greater than preset threshold.It, can be to the system to be accessed after obtaining the first model after training
Following calling amount is made a prediction.For example, providing No. 1-19 legacy system for 12306 systems and data being called to train mould
Type, then using model prediction 20 calling amounts after training, to predict the number of the user of No. 20 vote buyings using the calling amount
Amount.
Compared with prior art, the present embodiment calculates system to be accessed before the calling amount to system is predicted first
The system current status of system and each access system, if system to be accessed and each access system are not similar system,
The legacy system for treating access system calls data to carry out the pretreatment operation such as screening, and obtains stand-by system and calls data, so
Call the system features factor for obtaining the system to be accessed in data and element factor to be accessed as this from stand-by system afterwards
The factor data to be trained of system, using factor data training pattern to be trained to construct to obtain for predicting the system to be accessed
The model of following calling amount constructs prediction model using the method for big data analysis, to system to be accessed in this way
Following calling amount of system is accurately predicted, to cope with system exception problem;Unified pipe is carried out to the tune usage data of system
Reason is conducive to the accumulation and long-term analysis of data.
In a preferred embodiment, it in order to which the factor data to be trained for treating access system is adjusted in time, needs
Followed up and evaluated to prediction result, according to prediction result and actual conditions treat the factor data to be trained of access system into
Row evaluation and adjustment when above-mentioned processing system is executed by the processor, also realize following steps:
Factor data to be trained successively is rejected according to preset rejecting rule, is being rejected after training factor data every time,
The model is trained using factor data to be trained remaining after rejecting, for predicting the system to be accessed after being trained
Each second model of the calling amount of system;The calling amount for obtaining each second model prediction system to be accessed respectively is corresponding each
A accuracy rate evaluates the factor data to be trained rejected according to each accuracy rate.
Wherein, preset rejecting rule includes: that factor data to be trained sorts, and rejects some one by one according to sequencing
Factor data to be trained is rejected to training by two because of subnumber alternatively, sorting factor data to be trained according to sequencing
According to, etc..
Wherein, the second model and the first model are all RNN neural network model.
In a preferred embodiment, in order to preferably be managed important factor data to be trained, above-mentioned processing system
When system is executed by the processor, following steps are also realized:
If having the corresponding accuracy rate of the second model to be less than presets preset numerical value, will be rejected in training second model
Factor data to be trained recommended as important factor data;Establish the system to be accessed and corresponding important factor data
Incidence relation and preservation.
Wherein, if reject some or certain it is several after training factor data, the second model after corresponding training
Accuracy rate be less than presets preset numerical value, that is, be decreased obviously, then it is assumed that rejected some or certain it is several to training because of subnumber
According to for the data more important for the system to be accessed;If reject some or certain it is several after training factor data, it is corresponding
Training after the second model accuracy rate variation it is little, then it is assumed that some or certain the several factor datas to be trained rejected
For that can be excluded, not as the factor data to be trained of training pattern for the unessential data of system to be accessed.
For more important factor data to be trained, it can be recommended, it further, can also be to more important
Factor data to be trained establishes itself and the incidence relation of corresponding system to be accessed and preservation, with treat trained factor data into
The management of row system is conducive to the accumulation and long-term analysis of data.
As shown in Fig. 2, Fig. 2 is the flow diagram of one embodiment of method of forecasting system calling amount of the present invention, the prediction
The method of system calling amount the following steps are included:
Step S1 obtains the system parameter of system to be accessed and each access system, based on the system parameter respectively
Calculate the system current status of system to be accessed and each access system;
Whether step S12, the system current status for analyzing system to be accessed and access system are greater than default similarity threshold;
Wherein, server provides access interface to system to be accessed, receives the system parameter and system tune of system to be accessed
With data, it is respectively used to computing system similarity and training pattern.System parameter includes type of service, system architecture type, portion
Affix one's name to environmental form, portfolio and user volume.For example, type of service includes insurance, production danger, life insurance, vehicle insurance etc.;System architecture type
Including mobile terminal framework, PC end-rack structure etc.;Deployed environment type is the region of deployment, including goes up north wide deeply etc.;Portfolio and use
Family amount is number of levels.
In one embodiment, the system current status packet of system to be accessed and each access system is calculated based on system parameter
It includes: system to be accessed and each the similar value a1 of the type of service of access system, system is obtained based on the system parameter respectively
The similar value a2 of framework, the similar value a3 of deployed environment, the similar value a4 of portfolio and the similar value a5 of user volume, and obtain industry
The preset weight W1 of service type, the preset weight W2 of system architecture type, the preset weight W3 of deployed environment type, portfolio are pre-
If weight W4 and user volume preset weight W5, computing system similarity R=a1*W1+a2*W2+a3*W3+a4*W4+a5*
W5, wherein weight represents the system parameter for the significance level of system, if weight is bigger, the system parameter is more important.
In a specific example, for type of service, if the service class of system to be accessed and some access system
Type is all identical, then the corresponding similar value of access system has been 1 to the system to be accessed with this, if type of service have 3 kinds it is identical,
For example, system to be accessed and this access system have insurance, produce danger, 3 kinds of business of life insurance, then corresponding similar value be 0.8, such as
Fruit type of service have 2 kinds it is identical, for example, system to be accessed and this access system has insurance, produces danger, then corresponding similar value
It is 0.6.For system architecture, if system to be accessed is all identical as the system architecture of some access system, this is waiting
Entering system, the corresponding similar value of access system has been 1 with this, if system architecture have it is a kind of identical, for example, system to be accessed with
The system architecture type of the access system only includes mobile terminal framework, then corresponding similar value is 0.8.For deployed environment,
System to be accessed is all identical as the deployed environment of some access system, then access system is corresponding with this for the system to be accessed
Similar value be 1, if to have 3 identical for deployed environment, for example, the region of deployment all includes the broadly area that goes up north, then corresponding phase
It is 0.8 like value, if to have 2 identical for deployed environment, for example, the region of deployment all includes Beijing and Shanghai, then it is corresponding similar
Value is 0.6.For portfolio, if system to be accessed is with some, the number of levels of the portfolio of access system is identical, for example,
It is ten million rank, then the corresponding similar value of access system has been 1 to the system to be accessed with this, if the number of levels of portfolio
Not identical, for example, system to be accessed is ten million rank, access system is million ranks, then corresponding similar value is 0.6.It is right
In user volume, if system to be accessed is with some, the number of levels of the user volume of access system is identical, the system to be accessed
The corresponding similar value of access system has been 1 with this, if the number of levels of user volume is not identical, for example, system to be accessed is thousand
Ten thousand ranks, access system is million ranks, then corresponding similar value is 0.6.
Step S2 is obtained if the system current status for needing access system and access system is greater than default similarity threshold
The factor data for taking the access system, using the factor data as the factor data to be trained of the system to be accessed;
Wherein, default similarity threshold is, for example, 0.85.Access system is come using method identical with access system
Following calling amount is predicted, if there is the system current status of access system and the system to be accessed is greater than default similarity threshold
Factor data that access system uses can be directly used as the system to be accessed it may be considered that two systems are more similar in value
The factor data to be trained of system.
Wherein, factor data includes time, network type, host number, registration number of users, system calling amount, business
Type, festivals or holidays factor etc., which is used directly for training pattern.
Step S3, if the system current status of system to be accessed and each access system is respectively less than or is equal to default similarity threshold
Value then obtains the system of system to be accessed within a preset time and calls data, and the system for treating access system calls data to carry out
Pretreatment obtains stand-by system and calls data;
Wherein, if in access system, default phase is respectively less than or equal to the system current status of system to be accessed
Like degree threshold value, then it is assumed that the system to be accessed and each access system are dissimilar, therefore, it is desirable to obtain in system to be accessed
Legacy system calls data, by calling data to be pre-processed to obtain stand-by system calling data system.
In one embodiment, it includes: that acquisition is described to be accessed that the system for treating access system, which calls data to carry out pretreatment,
System calling amount of the system in the first preset time, calculates standard side of the system calling amount in first preset time
Difference;If the standard variance is 0, the system calling amount in first preset time is filtered, and obtain the system to be accessed and exist
System calling amount in second preset time, to calculate standard variance of the system calling amount in second preset time;If
The standard variance is not 0, then it is the data of sky and the data for not meeting predetermined format in data that filtration system, which is called, will be filtered
The system obtained afterwards calls data to call data as stand-by system.
Wherein, it is not pretreated data that system, which calls data, comprising: system calling amount, time, network type, master
Machine quantity, registration number of users, type of service, festivals or holidays factor etc., the first preset time is past a period of time, for example,
Past one month or two months, the second preset time was different from the first preset time, for example, the first preset time is the past
Two months, then the second preset time be past one month.It is within the past period by calculating system to be accessed
The standard variance for calling amount of uniting, the stability of the system to be accessed is determined by standard variance, i.e., whether abnormal conditions occurs.
If the standard variance of system to be accessed is 0, determine that system calling amount all within this time is equal, it cannot
As the factor data in limit of consideration.For example, for 12306 seat reservation systems, if the number bought tickets daily in one month in the past
Amount is equal, then is considered as abnormal situation, system in this one month call data cannot function as in limit of consideration because of subnumber
According to may filter that system in this one month calls data.In addition the system calling amount in the second preset time is taken, is to calculate this
Standard variance of the calling amount of uniting in second preset time, and analyze whether the standard variance is 0 again, if standard variance is
0 system filtered in second preset time calls data, then the system calling amount in other times section is taken to calculate standard side
Difference, and so on, until the system calling amount that can to obtain standard variance not be 0.
If the standard variance of system calling amount of the system to be accessed in the first preset time is not 0, for this
System in one preset time calls data to be handled, comprising: it is empty data in data that filtration system, which is called, and filtering is not inconsistent
The data of predetermined format are closed, for example, the data that date format does not meet predetermined format are filtered, value type is not predetermined number
The data of Value Types are filtered, and after above-mentioned filtration treatment, are obtained stand-by system and are called data.
Step S4 calls the system features factor of the data acquisition system to be accessed and basic based on the stand-by system
The factor, using the system features factor and element factor as the factor data to be trained of the system to be accessed;
Wherein, the system features factor and element factor that the system to be accessed is obtained in data are called from stand-by system,
Element factor is that substantially will affect the factor of the calling amount of each system to be accessed comprising time, network type, host
Quantity, registration number of users, system calling amount etc.;The system features factor is the distinctive influence system to be accessed of system to be accessed
The factor of the calling amount of system comprising type of service, festivals or holidays factor etc..The present embodiment be not limited to above system feature because
Son and element factor can select the system features factor and element factor according to the actual conditions of system to be accessed.
In a specific example, if system to be accessed be life insurance system, element factor include the time, network type,
It is host number, registration number of users, at least several in system calling amount, in the system features factor, if type of service includes
Life insurance, endowment insurance, then festivals or holidays factor is, for example, whether double 12 have whether the second of life insurance and endowment insurance kills the movable, same day Double Ninth Festival
Release life insurance and endowment insurance discount and discount number etc., these festivals or holidays factors largely influence system calling
Amount.
Step S5, the factor data to be trained based on system to be accessed is trained preset model, after being trained
For predicting the first model of the calling amount of the system to be accessed.
Wherein, preset model is preferably RNN neural network model.In one embodiment, above-mentioned system to be accessed is being utilized
During the factor data to be trained of system is trained RNN neural network model, it is first determined the RNN neural network model
Training number, if there is access system and the system to be accessed it is similar (for example, system current status be greater than it is a certain preset
Threshold value), then the corresponding frequency of training of model of access system can be used, if access system and the system to be accessed are all
Dissmilarity then calls the data volume of data to determine the number of training according to the stand-by system that system to be accessed provides, if to
System calls the data volume of data more, then trained number can be relatively smaller, if stand-by system calls data
Data volume is less, then trained number can be relatively more.
In one embodiment, RNN neural network model is carried out using the factor data to be trained of above-mentioned system to be accessed
Training includes: to collect remaining 20% or 25% as verifying using 75% data of factor data to be trained as training set,
After having trained RNN neural network model with training set, the RNN neural network model after training is tested using verifying collection
Card, if the accuracy rate of model is greater than preset threshold (such as 0.985), the RNN neural network model after training can be used for
Prediction can increase factor data to be trained and be trained again if accuracy rate is not more than preset threshold, until after training
The accuracy rate of RNN neural network model is greater than preset threshold.It, can be to the system to be accessed after obtaining the first model after training
Following calling amount is made a prediction.For example, providing No. 1-19 legacy system for 12306 systems and data being called to train mould
Type, then using model prediction 20 calling amounts after training, to predict the number of the user of No. 20 vote buyings using the calling amount
Amount.
Compared with prior art, the present embodiment calculates system to be accessed before the calling amount to system is predicted first
The system current status of system and each access system, if system to be accessed and each access system are not similar system,
The legacy system for treating access system calls data to carry out the pretreatment operation such as screening, and obtains stand-by system and calls data, so
Call the system features factor for obtaining the system to be accessed in data and element factor to be accessed as this from stand-by system afterwards
The factor data to be trained of system, using factor data training pattern to be trained to construct to obtain for predicting the system to be accessed
The model of following calling amount constructs prediction model using the method for big data analysis, to system to be accessed in this way
Following calling amount of system is accurately predicted, to cope with system exception problem;Unified pipe is carried out to the tune usage data of system
Reason is conducive to the accumulation and long-term analysis of data.
In a preferred embodiment, it in order to which the factor data to be trained for treating access system is adjusted in time, needs
Followed up and evaluated to prediction result, according to prediction result and actual conditions treat the factor data to be trained of access system into
Row evaluation and adjustment, as shown in figure 3, after the step S5, further includes:
Step S6 successively rejects factor data to be trained according to preset rejecting rule, is rejecting the factor to be trained every time
After data, the model is trained using factor data to be trained remaining after rejecting, for predicting this after being trained
Each second model of the calling amount of system to be accessed;
Step S7 obtains the corresponding each accuracy rate of calling amount of each second model prediction system to be accessed respectively,
The factor data to be trained rejected is evaluated according to each accuracy rate.
Wherein, preset rejecting rule includes: that factor data to be trained sorts, and rejects some one by one according to sequencing
Factor data to be trained is rejected to training by two because of subnumber alternatively, sorting factor data to be trained according to sequencing
According to, etc..
Wherein, the second model and the first model are all RNN neural network model.
In a preferred embodiment, in order to preferably be managed important factor data to be trained, step described above
After rapid S7, further includes:
If having the corresponding accuracy rate of the second model to be less than presets preset numerical value, will be rejected in training second model
Factor data to be trained recommended as important factor data;Establish the system to be accessed and corresponding important factor data
Incidence relation and preservation.
Wherein, if reject some or certain it is several after training factor data, the second model after corresponding training
Accuracy rate be less than presets preset numerical value, that is, be decreased obviously, then it is assumed that rejected some or certain it is several to training because of subnumber
According to for the data more important for the system to be accessed;If reject some or certain it is several after training factor data, it is corresponding
Training after the second model accuracy rate variation it is little, then it is assumed that some or certain the several factor datas to be trained rejected
For that can be excluded, not as the factor data to be trained of training pattern for the unessential data of system to be accessed.
For more important factor data to be trained, it can be recommended, it further, can also be to more important
Factor data to be trained establishes itself and the incidence relation of corresponding system to be accessed and preservation, with treat trained factor data into
The management of row system is conducive to the accumulation and long-term analysis of data.
The present invention also provides a kind of computer readable storage medium, processing is stored on the computer readable storage medium
The step of system, the processing system realizes the method for above-mentioned forecasting system calling amount when being executed by processor.
The serial number of the above embodiments of the invention is only for description, does not represent the advantages or disadvantages of the embodiments.
Through the above description of the embodiments, those skilled in the art can be understood that above-described embodiment side
Method can be realized by means of software and necessary general hardware platform, naturally it is also possible to by hardware, but in many cases
The former is more preferably embodiment.Based on this understanding, technical solution of the present invention substantially in other words does the prior art
The part contributed out can be embodied in the form of software products, which is stored in a storage medium
In (such as ROM/RAM, magnetic disk, CD), including some instructions are used so that a terminal device (can be mobile phone, computer, clothes
Business device, air conditioner or the network equipment etc.) execute method described in each embodiment of the present invention.
The above is only a preferred embodiment of the present invention, is not intended to limit the scope of the invention, all to utilize this hair
Equivalent structure or equivalent flow shift made by bright specification and accompanying drawing content is applied directly or indirectly in other relevant skills
Art field, is included within the scope of the present invention.
Claims (10)
1. a kind of server, which is characterized in that the server includes memory and the processor that connect with the memory, institute
The processing system for being stored with and being run in memory on the processor is stated, when the processing system is executed by the processor
Realize following steps:
The system parameter for obtaining system to be accessed and each access system respectively calculates system to be accessed based on the system parameter
With the respectively system current status of access system;
If the system current status for needing access system and access system is greater than default similarity threshold, obtains this and accessed and be
The factor data of system, using the factor data as the factor data to be trained of the system to be accessed, when the factor data includes
Between, network type, host number, registration number of users, system calling amount, type of service, festivals or holidays factor;
If the system current status of system to be accessed and each access system is respectively less than or is equal to default similarity threshold, obtain to
The system of access system within a preset time calls data, and the system for treating access system calls data to be pre-processed, and obtains
Stand-by system calls data;
Based on the system features factor and element factor of the stand-by system calling data acquisition system to be accessed, it is with this
The factor data to be trained of system characterization factor and element factor as the system to be accessed;
Factor data to be trained based on system to be accessed is trained preset model, after being trained for predict this to
First model of the calling amount of access system.
2. server according to claim 1, which is characterized in that the system parameter includes type of service, system architecture
Type, deployed environment type, portfolio and user volume, it is described to calculate system to be accessed based on the system parameter and respectively accessed
It the step of system current status of system, specifically includes:
System to be accessed and each the similar value a1 of the type of service of access system, system are obtained respectively based on the system parameter
The similar value a2 of framework, the similar value a3 of deployed environment, the similar value a4 of portfolio and the similar value a5 of user volume, and obtain industry
The preset weight W1 of service type, the preset weight W2 of system architecture type, the preset weight W3 of deployed environment type, portfolio are pre-
If weight W4 and user volume preset weight W5, computing system similarity R=a1*W1+a2*W2+a3*W3+a4*W4+a5*
W5。
3. server according to claim 1 or 2, which is characterized in that it includes system calling amount that the system, which calls data,
The system for treating access system calls data to carry out pretreated step, specifically includes:
Obtain system calling amount of the system to be accessed in the first preset time, calculate the system calling amount this first
Standard variance in preset time;
If the standard variance is 0, the system calling amount in first preset time is filtered, and obtain the system to be accessed and exist
System calling amount in second preset time, to calculate standard variance of the system calling amount in second preset time, directly
It is not 0 to the standard variance, first preset time and second preset time are time in the past, and described first is default
Time is not equal to second preset time;
If the standard variance is not 0, it is empty data and the data for not meeting predetermined format in data that filtration system, which is called,
Data are called to call data as the stand-by system system obtained after filtering.
4. server according to claim 1 or 2, which is characterized in that when the processing system is executed by the processor,
Also realize following steps:
Factor data to be trained successively is rejected according to preset rejecting rule, is rejecting after training factor data, is utilizing every time
Remaining factor data to be trained is trained the model after rejecting, for predicting the system to be accessed after being trained
Each second model of calling amount;
The corresponding each accuracy rate of calling amount for obtaining each second model prediction system to be accessed respectively, according to described each
Accuracy rate evaluates the factor data to be trained rejected.
5. a kind of method of forecasting system calling amount, which is characterized in that the method for the forecasting system calling amount includes:
S1 obtains the system parameter of system to be accessed and each access system respectively, is calculated based on the system parameter to be accessed
The system current status of system and each access system;
S2 obtains this and has connect if the system current status for needing access system and access system is greater than default similarity threshold
The factor data for entering system, using the factor data as the factor data to be trained of the system to be accessed, the factor data packet
Include time, network type, host number, registration number of users, system calling amount, type of service, festivals or holidays factor;
S3 is obtained if the system current status of system to be accessed and each access system is respectively less than or is equal to default similarity threshold
The system of system to be accessed within a preset time is taken to call data, the system for treating access system calls data to be pre-processed,
It obtains stand-by system and calls data;
S4 calls the system features factor and element factor of the data acquisition system to be accessed based on the stand-by system, with
The factor data to be trained of the system features factor and element factor as the system to be accessed;
S5, the factor data to be trained based on system to be accessed are trained preset model, for predicting after being trained
First model of the calling amount of the system to be accessed.
6. the method for forecasting system calling amount according to claim 5, which is characterized in that the system parameter includes business
Type, system architecture type, deployed environment type, portfolio and user volume, the step S1, specifically include:
System to be accessed and each the similar value a1 of the type of service of access system, system are obtained respectively based on the system parameter
The similar value a2 of framework, the similar value a3 of deployed environment, the similar value a4 of portfolio and the similar value a5 of user volume, and obtain industry
The preset weight W1 of service type, the preset weight W2 of system architecture type, the preset weight W3 of deployed environment type, portfolio are pre-
If weight W4 and user volume preset weight W5, computing system similarity R=a1*W1+a2*W2+a3*W3+a4*W4+a5*
W5。
7. the method for forecasting system calling amount according to claim 5 or 6, which is characterized in that the system calls data
Including system calling amount, the step S3 is specifically included:
Obtain system calling amount of the system to be accessed in the first preset time, calculate the system calling amount this first
Standard variance in preset time;
If the standard variance is 0, the system calling amount in first preset time is filtered, and obtain the system to be accessed and exist
System calling amount in second preset time, to calculate standard variance of the system calling amount in second preset time, directly
It is not 0 to the standard variance, first preset time and second preset time are time in the past, and described first is default
Time is not equal to second preset time;
If the standard variance is not 0, it is empty data and the data for not meeting predetermined format in data that filtration system, which is called,
Data are called to call data as stand-by system the system obtained after filtering.
8. the method for forecasting system calling amount according to claim 5 or 6, which is characterized in that after the step S5, also
Include:
S6 successively rejects factor data to be trained according to preset rejecting rule, is being rejected after training factor data every time, benefit
The model is trained with factor data to be trained remaining after rejecting, for predicting the system to be accessed after being trained
Calling amount each second model;
S7 obtains the corresponding each accuracy rate of calling amount of each second model prediction system to be accessed, according to described respectively
Each accuracy rate evaluates the factor data to be trained rejected.
9. the method for forecasting system calling amount according to claim 8, which is characterized in that after the step S7, also wrap
It includes:
If have the corresponding accuracy rate of the second model be less than preset preset numerical value, by training second model in rejected to
Training factor data is recommended as important factor data;
Establish the system to be accessed and the incidence relation of corresponding important factor data and preservation.
10. a kind of computer readable storage medium, which is characterized in that be stored with processing system on the computer readable storage medium
System, the forecasting system calling amount as described in any one of claim 5 to 9 is realized when the processing system is executed by processor
The step of method.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810554262.0A CN109039691B (en) | 2018-06-01 | 2018-06-01 | Server, method for predicting system call amount and storage medium |
PCT/CN2018/102091 WO2019227707A1 (en) | 2018-06-01 | 2018-08-24 | Server, method for predicting system invocation volume, and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810554262.0A CN109039691B (en) | 2018-06-01 | 2018-06-01 | Server, method for predicting system call amount and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109039691A true CN109039691A (en) | 2018-12-18 |
CN109039691B CN109039691B (en) | 2021-05-18 |
Family
ID=64611873
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810554262.0A Active CN109039691B (en) | 2018-06-01 | 2018-06-01 | Server, method for predicting system call amount and storage medium |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN109039691B (en) |
WO (1) | WO2019227707A1 (en) |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8473448B2 (en) * | 2007-08-22 | 2013-06-25 | Fujitsu Limited | Compound property prediction apparatus, property prediction method, and program for implementing the method |
CN103886391A (en) * | 2014-03-21 | 2014-06-25 | 广州杰赛科技股份有限公司 | Method and device for predicating service volume |
CN104581749A (en) * | 2013-10-11 | 2015-04-29 | 北京亿阳信通科技有限公司 | A method and device for predicting service amount of data service of mobile network |
CN105095230A (en) * | 2014-04-29 | 2015-11-25 | 国际商业机器公司 | Method and device for determining performance prediction model of data analysis application |
CN105224998A (en) * | 2015-09-08 | 2016-01-06 | 北京金山安全软件有限公司 | Data processing method and device for pre-estimation model |
CN106934514A (en) * | 2015-12-31 | 2017-07-07 | 华为技术有限公司 | The generation method and device of a kind of regression model |
CN107426759A (en) * | 2017-08-09 | 2017-12-01 | 广州杰赛科技股份有限公司 | The Forecasting Methodology and system of newly-increased base station data portfolio |
CN107832866A (en) * | 2017-09-26 | 2018-03-23 | 晶赞广告(上海)有限公司 | A kind of method for predicting and device, storage medium, terminal |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10558544B2 (en) * | 2011-02-14 | 2020-02-11 | International Business Machines Corporation | Multiple modeling paradigm for predictive analytics |
CN103152389A (en) * | 2013-02-01 | 2013-06-12 | 华南师范大学 | Method and system of responding peak access in cloud computer system |
CN107493205B (en) * | 2017-07-13 | 2020-08-14 | 华为技术有限公司 | Method and device for predicting capacity expansion performance of equipment cluster |
CN107885762B (en) * | 2017-09-19 | 2021-06-11 | 北京百度网讯科技有限公司 | Intelligent big data system, method and equipment for providing intelligent big data service |
-
2018
- 2018-06-01 CN CN201810554262.0A patent/CN109039691B/en active Active
- 2018-08-24 WO PCT/CN2018/102091 patent/WO2019227707A1/en active Application Filing
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8473448B2 (en) * | 2007-08-22 | 2013-06-25 | Fujitsu Limited | Compound property prediction apparatus, property prediction method, and program for implementing the method |
CN104581749A (en) * | 2013-10-11 | 2015-04-29 | 北京亿阳信通科技有限公司 | A method and device for predicting service amount of data service of mobile network |
CN103886391A (en) * | 2014-03-21 | 2014-06-25 | 广州杰赛科技股份有限公司 | Method and device for predicating service volume |
CN105095230A (en) * | 2014-04-29 | 2015-11-25 | 国际商业机器公司 | Method and device for determining performance prediction model of data analysis application |
CN105224998A (en) * | 2015-09-08 | 2016-01-06 | 北京金山安全软件有限公司 | Data processing method and device for pre-estimation model |
CN106934514A (en) * | 2015-12-31 | 2017-07-07 | 华为技术有限公司 | The generation method and device of a kind of regression model |
CN107426759A (en) * | 2017-08-09 | 2017-12-01 | 广州杰赛科技股份有限公司 | The Forecasting Methodology and system of newly-increased base station data portfolio |
CN107832866A (en) * | 2017-09-26 | 2018-03-23 | 晶赞广告(上海)有限公司 | A kind of method for predicting and device, storage medium, terminal |
Also Published As
Publication number | Publication date |
---|---|
CN109039691B (en) | 2021-05-18 |
WO2019227707A1 (en) | 2019-12-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103430151B (en) | A kind of resource allocation method and device | |
CN109120428B (en) | Method and system for wind control analysis | |
CN104050543B (en) | Event-handling method and stream processing system in stream processing system | |
CN110718057B (en) | Road network operation state evaluation method and device, electronic equipment and medium | |
CN111177129A (en) | Label system construction method, device, equipment and storage medium | |
CN110138688A (en) | Dynamic adjusts method, apparatus, equipment and the readable storage medium storing program for executing of business interface | |
CN109840111A (en) | A kind of patterned transaction processing system and method | |
CN110634021A (en) | Big data based vehicle estimation method, system, device and readable storage medium | |
CN107798341A (en) | User view Forecasting Methodology, electronic equipment and computer-readable recording medium | |
CN110445939B (en) | Capacity resource prediction method and device | |
CN109543891A (en) | Method for building up, equipment and the computer readable storage medium of capacity prediction model | |
CN107103453A (en) | Public emolument computational methods and system | |
CN110222535A (en) | Processing unit, method and the storage medium of block chain configuration file | |
CN108806302A (en) | A kind of vehicle dispatching method and device | |
CN107807967A (en) | Real-time recommendation method, electronic equipment and computer-readable recording medium | |
CN110533286A (en) | Allied member's management method, device and computer equipment based on block chain | |
CN115033340A (en) | Host selection method and related device | |
CN112835808A (en) | Interface testing method and device, computer equipment and storage medium | |
CN110188255B (en) | Power consumer behavior mining method and system based on business data sharing fusion | |
CN110060041A (en) | Channel of disbursement cut-in method, system, computer equipment and readable storage medium storing program for executing | |
CN109413166A (en) | A kind of industrial gateway and its data managing method | |
CN106201490A (en) | The method and device that a kind of optimization application is run | |
CN109039691A (en) | The method and storage medium of server, forecasting system calling amount | |
CN114841797A (en) | Method and device for determining business processing rule based on Drools rule engine | |
CN112637793B (en) | Scene charging method, system, electronic equipment and storage medium based on 5G |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |