CN108460490A

CN108460490A - A kind of prediction technique, device and the equipment of business occurrence quantity

Info

Publication number: CN108460490A
Application number: CN201810220193.XA
Authority: CN
Inventors: 黄馨誉; 吴蔚川
Original assignee: Alibaba Group Holding Ltd
Current assignee: Advanced New Technologies Co Ltd; Advantageous New Technologies Co Ltd
Priority date: 2018-03-16
Filing date: 2018-03-16
Publication date: 2018-08-28
Also published as: WO2019174410A1; TWI703461B; TW201939315A

Abstract

This specification embodiment discloses a kind of prediction technique, device and the equipment of business occurrence quantity, and this method includes：History service data before predetermined amount of time are subjected to sliding-model control, obtain the business occurrence quantity vector of time granularity, in addition, it can be according to the business occurrence quantity of continuous time in history service data, generation business occurrence quantity distribution characteristics vector, finally, the business occurrence quantity in predetermined amount of time can be determined according to the business occurrence quantity vector sum business occurrence quantity distribution characteristics vector of time granularity.

Description

A kind of prediction technique, device and the equipment of business occurrence quantity

Technical field

This specification is related to field of computer technology more particularly to a kind of prediction technique of business occurrence quantity, device and sets It is standby.

Background technology

With the continuous development of network technology and terminal technology, e-commerce is more and more important in people's daily life, For example, people can buy extensive stock etc. by network payment in shopping website.Moreover, under the outsourcing of line Shanghai and line The business such as pay face to face also to be grown rapidly, in this way, payment using (such as Alipay) need to support between trade company, buyer with Different currency carries out Zhi Fuyu gatherings.In this way, payment application just need settle accounts respective country currency give corresponding businessman, because This, there is demand of largely exchanging in payment application.

In general, payment application needs to buy a certain number of foreign exchanges to cope with business demand, in order to the greatest extent in the every workday Influence of the amount reduction fluctuation of exchange rate to payment application, the same day locks buying foreign exchange for the same day with counterparty on weekdays for payment application meeting The amount of money.It in practical applications, can be specific such as the method for moving average, sliding average, ARIMA by time series algorithm (Autoregressive Integrated Moving Average Model, autoregression integrate moving average model) or Holt-Winters etc. provides the development trend of predetermined amount of time, and still, above-mentioned time series algorithm is to time series trend Coherence request is higher, if nearest business development trend has exception, the prediction result obtained according to above-mentioned algorithm has very much It may be also abnormal, cause predicted value deviation larger, in this way, it is necessary to which business can accurately be predicted in real time by providing one kind Amount, and business risk can be reduced and improve the scheme of fund utilization ratio.

Invention content

The purpose of this specification embodiment is to provide a kind of prediction technique, device and the equipment of business occurrence quantity, to provide One kind can accurately predict business occurrence quantity in real time, meanwhile, it is capable to reduce business risk and improve the side of fund utilization ratio Case.

To realize that above-mentioned technical proposal, this specification embodiment are realized in：

A kind of prediction technique for business occurrence quantity that this specification embodiment provides, the method includes：

History service data before predetermined amount of time are subjected to sliding-model control, the business for obtaining time granularity occurs Amount vector, and according to the business occurrence quantity of continuous time in the history service data, generate business occurrence quantity distribution characteristics Vector；

Business occurrence quantity distribution characteristics vector, determines institute described in business occurrence quantity vector sum according to the time granularity State the business occurrence quantity in predetermined amount of time.

Optionally, the history service data include the first of the first time period nearest apart from the predetermined amount of time going through History business datum and the second history service data in addition to the first history service data,

Continuous time in history service data described in the business occurrence quantity vector sum according to the time granularity Business occurrence quantity generates business occurrence quantity distribution characteristics vector, including：

According to the business occurrence quantity of continuous time in the first history service data, the of business occurrence quantity distribution is generated One feature vector；

According to the business occurrence quantity of continuous time in the second history service data, the of business occurrence quantity distribution is generated Two feature vectors.

Optionally, business occurrence quantity distribution characteristics described in the business occurrence quantity vector sum according to the time granularity Vector, before determining the business occurrence quantity in the predetermined amount of time, the method further includes：

Determine the similarity between the first eigenvector and the second feature vector；

According to the similarity between the first eigenvector and the second feature vector, determine the second feature to The weight of amount.

Optionally, the similarity between the determination first eigenvector and the second feature vector, including：

The similarity between the first eigenvector and the second feature vector is determined by following any method： The included angle cosine value of Euclidean distance, vector, and vectorial absolute value of the difference.

Optionally, business occurrence quantity distribution characteristics described in the business occurrence quantity vector sum according to the time granularity Vector determines the business occurrence quantity in the predetermined amount of time, including：

Respectively by the business occurrence quantity vector of the time granularity and the first eigenvector and the second feature Vector merges, the first eigenvector after being merged and second feature vector；

It is excellent by loss function and predefined parameter based on the weight of second feature vector described in the second feature vector sum Change algorithm to optimize initial parameter, the initial parameter after being optimized；

According to after optimization initial parameter and the first eigenvector, determine that business in the predetermined amount of time occurs Amount.

Optionally, the predefined parameter optimization algorithm includes gradient descent algorithm, Newton method, quasi-Newton method, conjugate gradient Method and heuristic value.

A kind of prediction meanss for business occurrence quantity that this specification embodiment provides, described device include：

Processing module obtains the time for the history service data before predetermined amount of time to be carried out sliding-model control The business occurrence quantity vector of granularity, and according to the business occurrence quantity of continuous time in the history service data, generate business Occurrence quantity distribution characteristics vector；

Business prediction of emergence size module, for business hair described in the business occurrence quantity vector sum according to the time granularity Raw amount distribution characteristics vector, determines the business occurrence quantity in the predetermined amount of time.

The processing module, including：

First eigenvector generation unit, for being occurred according to the business of continuous time in the first history service data Amount generates the first eigenvector of business occurrence quantity distribution；

Second feature vector generation unit, for being occurred according to the business of continuous time in the second history service data Amount generates the second feature vector of business occurrence quantity distribution.

Optionally, described device further includes：

Similarity determining module, it is similar between the first eigenvector and the second feature vector for determining Degree；

Weight determination module is used for according to the similarity between the first eigenvector and the second feature vector, Determine the weight of the second feature vector.

Optionally, the similarity determining module determines the first eigenvector for passing through following any device Similarity between the second feature vector：The included angle cosine value of Euclidean distance, vector, and the difference of vector are absolute Value.

Optionally, the business prediction of emergence size module, including：

Combining unit, for respectively by the business occurrence quantity vector of the time granularity and the first eigenvector and The second feature vector merges, the first eigenvector after being merged and second feature vector；

Initial parameter optimizes unit, for based on second feature vector, being optimized by loss function and predefined parameter Algorithm optimizes initial parameter, the initial parameter after being optimized；

Business prediction of emergence size unit, for according to after optimization initial parameter and the first eigenvector, determine institute State the business occurrence quantity in predetermined amount of time.

A kind of pre- measurement equipment for business occurrence quantity that this specification embodiment provides, the pre- measurement equipment of the business occurrence quantity Including：

Processor；And

It is arranged to the memory of storage computer executable instructions, the executable instruction makes the place when executed Manage device：

The technical solution that is there is provided by above this specification embodiment as it can be seen that this specification embodiment by by predetermined amount of time History service data before carry out sliding-model control, the business occurrence quantity vector of time granularity are obtained, furthermore it is also possible to root According to the business occurrence quantity of continuous time in history service data, business occurrence quantity distribution characteristics vector is generated, it finally, can basis The business occurrence quantity vector sum business occurrence quantity distribution characteristics vector of time granularity determines that the business in predetermined amount of time occurs Amount, in this way, by business occurrence quantity distribution characteristics vectorial (may include multidimensional characteristic vectors) come to the industry in predetermined amount of time Business occurrence quantity is predicted, final prediction result can be accordingly adjusted in the case where short period service occurrence quantity fluctuates larger, Effectively business occurrence quantity is avoided to occur in the case where short-term fluctuation larger impact final prediction result, moreover, being sent out by business The mode of raw amount distribution characteristics vector, can also preferably capture the variation tendency of business occurrence quantity in designated time period (such as The variation tendency etc. of business occurrence quantity in history service data), so as to improve the predictablity rate of business occurrence quantity, reduce Business risk and improve fund utilization ratio.

Description of the drawings

In order to illustrate more clearly of this specification embodiment or technical solution in the prior art, below will to embodiment or Attached drawing needed to be used in the description of the prior art is briefly described, it should be apparent that, the accompanying drawings in the following description is only Some embodiments described in this specification, for those of ordinary skill in the art, in not making the creative labor property Under the premise of, other drawings may also be obtained based on these drawings.

Fig. 1 is a kind of prediction technique embodiment of business occurrence quantity of this specification；

Fig. 2 is a kind of structural schematic diagram of the forecasting system of business occurrence quantity of this specification；

Fig. 3 is the prediction technique embodiment of this specification another kind business occurrence quantity；

Fig. 4 is a kind of prediction meanss embodiment of business occurrence quantity of this specification；

Fig. 5 is a kind of prediction apparatus embodiments of business occurrence quantity of this specification.

Specific implementation mode

This specification embodiment provides a kind of prediction technique, device and the equipment of business occurrence quantity.

In order to make those skilled in the art more fully understand the technical solution in this specification, below in conjunction with this explanation Attached drawing in book embodiment is clearly and completely described the technical solution in this specification embodiment, it is clear that described Embodiment be only this specification a part of the embodiment, instead of all the embodiments.The embodiment of base in this manual, The every other embodiment that those of ordinary skill in the art are obtained without creative efforts, should all belong to The range of this specification protection.

Embodiment one

As shown in Figure 1, this specification embodiment provides a kind of prediction technique of business occurrence quantity, the executive agent of this method Can be terminal device or server, wherein the terminal device can such as personal computer equipment, can also be such as mobile phone, flat The mobile terminal devices such as plate computer, the terminal device can be the terminal device that user uses.The server can be independent Server can also be the server cluster etc. being made of multiple servers.This method can be used for accurately prediction industry in real time It is engaged in the processing such as occurrence quantity, is illustrated by taking server as an example in the present embodiment, it, can be under the case where for terminal device Related content processing is stated, details are not described herein.This method can specifically include following steps：

In step s 102, the history service data before predetermined amount of time are subjected to sliding-model control, obtain the time The business occurrence quantity vector of granularity.

Wherein, predetermined amount of time may be set according to actual conditions, specifically if predetermined amount of time can be the certain of future Duration, such as from current point in time to interval 1 hour after time point, can also be before current point in time sometime Section etc..Sliding-model control can be that limited individual in infinite space is mapped in limited space, be calculated with this to improve The processing mode of the spatiotemporal efficiency of method that is to say that sliding-model control is under conditions of not change data relative size, to data The processing mode reduced accordingly, sliding-model control are very big in data itself, its own can not be as the subscript of array Corresponding attribute is preserved, if only needing the relative priority of the data at this time, sliding-model control can be carried out to the data, That is to say when data only it is related with the relative size between them, and it is unrelated with the particular content of data when, can be to the data Carry out sliding-model control.Business occurrence quantity vector can be the related data that user generates when completing certain one or more business Vector.Business occurrence quantity can be number of users or the user's progress or complete for carrying out or completing certain one or more business At business quantity etc., for example, as shown in Fig. 2, wherein may include multiple servers, different servers can be difference Business service is provided, people can complete different business by different service servers, and business occurrence quantity may include It is a variety of, such as trading volume, inversion quantity etc..Time granularity can be that business occurs or the basic time of business occurrence quantity statistics is single Position, such as 10 minutes or 15 minutes.

In force, due to the outsourcing of line Shanghai, the fast development of business is paid under line face to face, (such as Alipay) is applied in payment It supports to carry out Zhi Fuyu gatherings between trade company, buyer with different currency.If trade company and buyer (i.e. user) are belonging respectively to not Same country, then either on line or under line, user only needs to pay using the currency locally used, this Sample, payment application just need to settle accounts corresponding foreign currency to different businessmans, and therefore, payment application has demand of largely exchanging, In this way, payment application needs to buy a certain number of foreign exchanges in the every workday to cope with business demand.It is converged to reduce to the greatest extent Rate fluctuates the influence to paying application, locks the amount bought on the same day on the day of payment application meeting on weekdays with counterparty, this Sample, payment application needs to buy a certain number of foreign exchanges in the every workday to cope with business demand, therefore needs are a kind of accurately Real-time method can predict business occurrence quantity, reduce business risk and improve fund utilization ratio.It in practical applications, can be with By time series algorithm, the specific such as method of moving average, sliding average, ARIMA or Holt-Winters, entering ginseng is Time series numerical value, and provide according to different algorithms the development trend of predetermined amount of time, still, above-mentioned time series algorithm pair The coherence request of time series trend is higher, i.e., if nearest business development trend has exception, is obtained according to above-mentioned algorithm The prediction result gone out is also probably abnormal, and above-mentioned time series algorithm is not high for the utilization rate of real time data, The variation tendency that the same day newest business occurrence quantity can not accurately be captured in practice, to cause predicted value deviation larger.

Furthermore it is also possible to predict that business yield, this method pass through business yield per hour in real time by rule of three Shared ratio calculates the total business yield on the same day in conjunction with the business yield in this real-time hour.And this method pair The coherence request of the ratio shared by business yield is higher per hour, if cataclysm occurs for the business yield of this hour, It is likely to be abnormal according to the prediction result that this method obtains.

As it can be seen that in order to accurately capture the variation tendency of the same day newest business occurrence quantity, reduce prediction error, it can be with It handles in the following manner, can specifically include the following contents：

It can first determine and need period for predicting, such as the time point from current point in time to after being spaced 2 hours is specifically such as Current point in time was 10 o'clock, and time point of the interval after 2 hours was 12 o'clock, then needed (i.e. predetermined time period predicted Section) can be 10 o'clock~12 o'clock；For another example, current point in time was 10 o'clock, needed (i.e. predetermined time period predicted Section) can be 24 o'clock etc. on 10 o'clock~same day.The related data of certain one or more business can be stored in server, The data can be that user is asking relevant business, and the related data generated after completing this business.Server determines It, can be from the corresponding service extracting data of storage in above-mentioned predetermined amount of time (10 points of such as current time after predetermined amount of time 24 o'clock on clock~same day) before certain time length (such as the previous moon or two months at 10 o'clock etc. at current time) history Business datum.In general, the data volume of the history service data is larger, it, can be to obtaining in order to reduce the processing pressure of server History service data carry out sliding-model control.It, can be by history service number by the sliding-model control to history service data According to the business occurrence quantity vector according to scheduled regular discrete at time granularity, wherein history service data can be divided For two parts, a part therein can be that occurs for newest business (i.e. the moment from the zero to real time traffic data on the same day Current point in time) history service data, the partial history business datum can according to it is above-mentioned rule by it is discrete be business Amount vector；Another part can also be divided into two for the history service data before the zero on the same day, the partial data A part, one of part can be history services at the time of current point in time corresponds in each day zero to this day Data, for example, current point in time was 10 o'clock, the date on the same day is 28 days 2 months, then the history service data of the part can wrap Include the history service data at 27 days 2 months zero clock~10 o'clock, the history service data at 26 days 2 months zero clock~10 o'clock, 2 months The history service data at zero clock on the 25th~10 o'clock, and so on；Another part can be current point in time pair in each day At the time of answering~the history service data at 24 o'clock of this day, for example, be based on above-mentioned example, the history service data of the part can be with History service data, the history service data at 2 months 26 10 o'clock~24 o'clock including 2 months 27 10 o'clock~24 o'clock, 2 The history service data at 25 10 o'clock~24 o'clock moon, and so on.The history service data of each above-mentioned part can be with According to above-mentioned rule by discrete for business occurrence quantity vector.

It should be noted that can be by the industry of the same day remaining moment (i.e. 24 o'clock on the 10 o'clock~same day at current time) The total business occurrence quantity of the sum of business occurrence quantity, that is to say the business occurrence quantity that needs are predicted.

In step S104, according to the business occurrence quantity of continuous time in above-mentioned history service data, business is generated Measure distribution characteristics vector.

In force, the rule or algorithm for generating feature vector, corresponding rule or calculation can be preset in server Method can determines according to actual conditions, and this specification embodiment does not limit this.Server can be first from obtained history industry It is engaged in the business occurrence quantity of extracting data continuous time, it can will be in history service data by the rule or algorithm of above-mentioned setting The business occurrence quantity of continuous time generates individual features to characterize the distribution situation of business occurrence quantity, the feature of above-mentioned generation Think distribution characteristics, then, the time that server can will obtain in distribution characteristics obtained above and above-mentioned steps S102 The business occurrence quantity vector of granularity merges, and obtains business occurrence quantity distribution characteristics vector.Wherein, the distribution of business occurrence quantity is special Levying vector can be there are many form of expression, such as average traffic occurrence quantity, the per hour amplification of business occurrence quantity and every per hour The rate of rise etc. of hour business occurrence quantity.

In practical applications, can also include that other are special other than above-mentioned common business occurrence quantity distribution characteristics vector Vector is levied, such as (specific such as business occurrence quantity distribution characteristics per week is vectorial, every for periodic traffic occurrence quantity distribution characteristics vector Season business occurrence quantity distribution characteristics vector, each month business occurrence quantity distribution characteristics vector), the distribution of live traffic occurrence quantity Feature vector (the specific business occurrence quantity distribution characteristics such as promotion day is vectorial), sliding rolling average business occurrence quantity distribution characteristics Vector (specifically such as calculates the average traffic occurrence quantity distribution characteristics vector in certain window some cycles).

In step s 106, according to the business occurrence quantity vector sum business occurrence quantity distribution characteristics of above-mentioned time granularity to Amount determines the business occurrence quantity in predetermined amount of time.

In force, really according to the business occurrence quantity vector sum business occurrence quantity distribution characteristics vector of above-mentioned time granularity Determining the concrete processing procedure of the business occurrence quantity in predetermined amount of time can be accomplished in several ways, in order to can be very good from It is fitted in the business occurrence quantity trend (or being properly termed as passing trend) that history service data are embodied and best suits certain for the moment Between the interior business occurrence quantity trend of section (i.e. predetermined amount of time) distribution, may be used dictionary learning algorithm realize to predetermined amount of time The prediction of interior business occurrence quantity, wherein dictionary learning algorithm may include two stages, i.e., dictionary builds the stage and utilizes word Allusion quotation indicates that business occurrence quantity stage, each stage in the above-mentioned two stage can be realized by many algorithms of different.Dictionary It is substantially that a kind of dimensionality reduction of huge data set is indicated to practise, in addition, dictionary learning always contain in history industry by trial learning Most simple feature in data deep layer of being engaged in.Dictionary learning can be that business occurrence quantity used in prediction business occurrence quantity is distributed The linear expression of feature vector and initial parameter, wherein initial parameter can be based on history service data by loss function and Parameter optimization algorithm carries out continuous iteration optimization and obtains, and loss function therein can be used for measuring above-mentioned business occurrence quantity point The fitting degree of the linear function of cloth feature vector and initial parameter can be based on existing linear function and calculate the generation of prediction business The penalty values and gradient of amount and the actual business occurrence quantity of history then mean to be fitted journey when obtained loss function minimum Spend optimal, corresponding initial parameter is optimized parameter, and the minimum value of loss function can be asked by parameter optimization algorithm Solution.Parameter optimization algorithm therein may include a variety of, for example, can be optimal to get with specifically used gradient descent method Parameter.It is possible, firstly, to which (numerical value of the gradient can be calculated the gradient of the loss function of determining current location by loss function To), initial step length can be used to be multiplied by the gradient of loss function, obtain the distance of current location decline, at this point it is possible to judge be No the distance that gradient declines is both less than the error amount set for all coefficients, if the distance that gradient declines is less than The error amount of setting, then terminal parameter optimization algorithm, initial parameter has been optimized parameter at this time, if the distance that gradient declines is big In the error amount of setting, then can update all initial parameters, and continue above-mentioned iterative calculation, until gradient decline away from Until the error amount less than setting.

Server can send out the business of time granularity obtained above by preset processing rule or algorithm Raw amount vector sum business occurrence quantity distribution characteristics vector merges processing, obtain with predict the relevant multidimensional of business occurrence quantity to Amount combination, to which the business occurrence quantity distribution characteristics after being merged is vectorial, the business occurrence quantity distribution characteristics vector after merging Can be and the relevant feature vector of the prediction of business occurrence quantity.Server can be based on history service data respectively to above-mentioned damage It loses function and parameter optimization algorithm is trained, the relevant parameter in loss function and parameter optimization algorithm is determined, to obtain Loss function after training and parameter optimization algorithm, it is then possible to based on loss function and parameter optimization algorithm pair after training Initial parameter carries out continuous iteration optimization, obtains optimized parameter.It is then possible to use above-mentioned business occurrence quantity distribution characteristics vector (the corresponding business occurrence quantity of history service data at moment can occur with newest business in the zero to real time traffic data on the same day Based on distribution characteristics vector) and optimized parameter obtained above progress linear regression, the business in predetermined amount of time can be obtained Occurrence quantity.

It should be noted that in practical applications, not only to the processing of the prediction of the business occurrence quantity in predetermined amount of time Can only be realized, can also be realized by other algorithms by dictionary learning algorithm, for example, sparse model related algorithm etc., this Specification embodiment does not limit this.

This specification embodiment provides a kind of prediction technique of business occurrence quantity, by by the history before predetermined amount of time Business datum carries out sliding-model control, the business occurrence quantity vector of time granularity is obtained, furthermore it is also possible to according to history service The business occurrence quantity of continuous time in data generates business occurrence quantity distribution characteristics vector, finally, can be according to time granularity Business occurrence quantity vector sum business occurrence quantity distribution characteristics vector determine the business occurrence quantity in predetermined amount of time, in this way, logical Cross business occurrence quantity distribution characteristics vectorial (may include multidimensional characteristic vectors) come to the business occurrence quantity in predetermined amount of time into Row prediction, final prediction result can be accordingly adjusted in the case where short period service occurrence quantity fluctuates larger, effectively avoids industry Business occurrence quantity occurs in the case where short-term fluctuation larger impact final prediction result, moreover, being distributed by business occurrence quantity special The mode for levying vector, can also preferably capture variation tendency (such as history service number of business occurrence quantity in designated time period According to the variation tendency etc. of middle business occurrence quantity), so as to improve the predictablity rate of business occurrence quantity, reduce business risk with And improve fund utilization ratio.

Embodiment two

As shown in figure 3, this specification embodiment provides a kind of prediction technique of business occurrence quantity, the executive agent of this method Can be terminal device or server, wherein the terminal device can such as personal computer equipment, can also be such as mobile phone, flat The mobile terminal devices such as plate computer, the terminal device can be the terminal device that user uses.The server can be independent Server can also be the server cluster etc. being made of multiple servers.This method can be used for accurately prediction industry in real time It is engaged in the processing such as occurrence quantity, is illustrated by taking server as an example in the present embodiment, it, can be under the case where for terminal device Related content processing is stated, details are not described herein.This method can specifically include following steps：

In step s 302, the history service data before predetermined amount of time are subjected to sliding-model control, obtain the time The business occurrence quantity vector of granularity.

In force, server can record the associated traffic data that every business generates, the related service of record in real time May include two parts in data, a part is the business datum generated on the same day (i.e. from the zero on the same day to current time), separately A part can be the same day before history service data.Server can obtain above-mentioned two-part business datum, can be right Above-mentioned two-part business datum carries out data cleansing according to certain rule.It can be to above-mentioned two-part by data cleansing Business datum examine and verify again, to which the duplicate message in two-part business datum be deleted, and can be to it Present in mistake correct etc..After business datum after being corrected, sliding-model control can be carried out to it, when obtaining Between granularity business occurrence quantity vector, wherein carrying out sliding-model control to history service data obtains the business of time granularity The concrete processing procedure of occurrence quantity vector may refer to the related content in above-described embodiment one in step S102, no longer superfluous herein It states.

May include two parts history service data based on the above, in server, therefore, history service data can be with The first history service data including preset distance period nearest first time period and remove the first history service data The second outer history service data, preset distance period therein nearest first time period can refer to that same day zero starts To the period of current point in time, the second history service data can be the same day before history service data, and the second history Business datum can also include two parts, and one of part can be current point in time pair in each day zero to the day History service data at the time of answering, another part can be at the time of current point in time corresponds in each day to 24 o'clock of day History service data.For the history service data of the multiple portions of above-mentioned division can execute respectively following steps S304~ The processing of step S316.

In step s 304, according to the business occurrence quantity of continuous time in above-mentioned first history service data, business is generated The first eigenvector of occurrence quantity distribution.

In force, the rule or algorithm for generating feature vector, corresponding rule or calculation can be preset in server Method can determines according to actual conditions, and this specification embodiment does not limit this.Server can first be gone through from first obtained The business occurrence quantity of continuous time is extracted in history business datum, it can be by the rule of above-mentioned setting or algorithm by the first history industry The business occurrence quantity of continuous time generates the first eigenvector of business occurrence quantity distribution in data of being engaged in.For example, if first goes through History business datum be same day zero start the history service data to current point in time, then based on the first history service data when Between granularity business occurrence quantity vector, the first eigenvector of 1 × N-dimensional can be obtained.

In step S306, according to the business occurrence quantity of continuous time in above-mentioned second history service data, business is generated The second feature vector of occurrence quantity distribution.

In force, server can first occur from the business of the second obtained history service extracting data continuous time The business occurrence quantity of continuous time in second history service data can be generated industry by amount by the rule or algorithm of above-mentioned setting The second feature that occurrence quantity is distributed of being engaged in is vectorial.For example, to each day zero in the second history service data start into the day when History service data at the time of preceding time point corresponds to, then the business occurrence quantity of the time granularity based on the history service data Vector can obtain the second feature vector of 1 × N-dimensional；Current point in time in each day in second history service data is corresponded to At the time of start the history service data to 24 o'clock of this day, then the time granularity based on the history service data business hair Raw amount vector can obtain the second feature vector of 1 × N-dimensional.

In step S308, respectively by the business occurrence quantity vector of above-mentioned time granularity and above-mentioned first eigenvector and Above-mentioned second feature vector merges, the first eigenvector after being merged and second feature vector.

In force, server can by first eigenvector obtained above and second feature vector respectively with above-mentioned step The business occurrence quantity vector of the time granularity obtained in rapid S302 merges, and the business occurrence quantity distribution after being merged is special Sign vector (first eigenvector after merging and second feature vector).Wherein, first eigenvector and second feature vector Can be there are many form of expression, such as average traffic occurrence quantity, the per hour amplification of business occurrence quantity and industry per hour per hour The rate of rise etc. of business occurrence quantity.

It should be noted that the first eigenvector and second feature vector after above-mentioned merging may include as put down per hour Equal business occurrence quantity, the per hour amplification of business occurrence quantity and the per hour rate of rise etc. of business occurrence quantity, further, it is also possible to Including such as periodic traffic occurrence quantity distribution characteristics vector, live traffic occurrence quantity distribution characteristics vector, sliding rolling average industry Business occurrence quantity distribution characteristics vector etc..

Can be first eigenvector and second feature vector point to show the important relationship of each different characteristic vector Corresponding weight is not set, the processing of following step S310 and step S312 are specifically may refer to.

In step S310, the similarity between above-mentioned first eigenvector and above-mentioned second feature vector is determined.

In force, it can judge the two weight by the similarity between first eigenvector and second feature vector Numerical value, the related algorithm of similarity calculation can be set for thus, in practical applications, following any side can be passed through Method determines the similarity between first eigenvector and second feature vector：The included angle cosine value of Euclidean distance, vector, Yi Jixiang The absolute value of the difference of amount.By taking Euclidean distance as an example, server can calculate between first eigenvector and second feature vector Euclidean distance, by the data of obtained Euclidean distance, it is possible to determine that the phase between first eigenvector and second feature vector Like degree, i.e., when the Euclidean distance between first eigenvector and second feature vector is smaller, then show first eigenvector and Closer between two feature vectors, the similarity of the two is higher, subsequently carry out the business prediction of emergence size when, the second feature to The proportion of amount is bigger；When the Euclidean distance between first eigenvector and second feature vector is bigger, then show fisrt feature to It measures and differs bigger (or difference is bigger) between second feature vector, the similarity of the two is lower, is subsequently carrying out business When amount prediction, the proportion of second feature vector is smaller.Correspondingly, for the exhausted of the difference of vectorial included angle cosine value or vector To be worth be similarity calculation algorithm specific processing mode and judgment mode, can be with above-mentioned Euclidean distance the case where, is similar, tool Body may refer to the processing procedure of Euclidean distance, and details are not described herein.

In step S312, according to the similarity between first eigenvector and second feature vector, second feature is determined The weight of vector.

In force, the processing of server S310 through the above steps obtains first eigenvector and second feature vector Between similarity after, the weight of second feature vector can be determined based on the similarity of the two, in practical applications can will be upper The numerical value of the Euclidean distance between the first eigenvector being calculated and second feature vector is stated as second feature vector Weight, alternatively, can be using the included angle cosine value between first eigenvector and second feature vector as second feature vector Weight, alternatively, can be using the absolute value of the difference of first eigenvector and second feature vector as the weight of second feature vector Deng.

The business occurrence quantity in predetermined amount of time can be determined by dictionary learning algorithm, in dictionary learning algorithm therein May include initial parameter, moreover, loss function and parameter optimization algorithm are also related in dictionary learning algorithm, it is specific processed Journey may refer to the processing of following step S314 and step S316.

In step S314, based on the weight of above-mentioned second feature vector sum second feature vector, by loss function and Predefined parameter optimization algorithm optimizes initial parameter, the initial parameter after being optimized.

Wherein, predefined parameter optimization algorithm may include a variety of, and the algorithm of plurality of optional presented below may include gradient Descent algorithm, Newton method, quasi-Newton method, conjugate gradient method and heuristic value etc..Loss function can refer to a kind of incites somebody to action One event is mapped to an expression and the function on the real number of the relevant economic cost of the event or opportunity cost.

In force, the particular content of loss function and parameter optimization algorithm can be preset in the server, such as Gradient descent algorithm or Newton method can be set as parameter optimization algorithm, and can be by the specific of preset loss function Form is written in server.Server can pass through the corresponding second feature vector sum second feature of the second history service data The weight and preset loss function and parameter optimization algorithm of vector, to the initial parameter in dictionary learning algorithm into Row iteration calculates, and can obtain optimal initial parameter by multiple iterative calculation, and then the initial ginseng after being optimized Number.

In step S316, according to after optimization initial parameter and above-mentioned first eigenvector, determine in predetermined amount of time Business occurrence quantity.

The concrete processing procedure of above-mentioned steps S316 may refer to the related content in step S106 in above-described embodiment one, Details are not described herein.

The processing procedure of S302~step S316 through the above steps, is predicted by using the mode of multidimensional characteristic vectors Business occurrence quantity, it is not high to the coherence request of time trend in the processing procedure, such as in the recent period or close on business in 1 hour Occurrence quantity fluctuation is larger (alternatively, go up or decline and is apparent), then the processing procedure can automatically adjust finally obtained prediction knot Fruit；In addition, the processing procedure can preferably capture the variation tendency of business occurrence quantity in designated time period, and pass through dictionary Learning algorithm can be very good the fitting from the variation tendency of business occurrence quantity in history service data and best suit real time data Trend is distributed.It is accurate better than being predicted by time series approach to be obtained in real-time prediction using the processing procedure Rate.

Embodiment three

It is the prediction technique for the business occurrence quantity that this specification embodiment provides above, is based on same thinking, this explanation Book embodiment also provides a kind of prediction meanss of business occurrence quantity, as shown in Figure 4.

The prediction meanss of the business occurrence quantity include：Processing module 401 and business prediction of emergence size module 402, wherein：

Processing module 401 obtains the time for the history service data before predetermined amount of time to be carried out sliding-model control The business occurrence quantity vector of granularity, and according to the business occurrence quantity of continuous time in the history service data, generate industry Occurrence quantity distribution characteristics of being engaged in is vectorial；

Business prediction of emergence size module 402, for industry described in the business occurrence quantity vector sum according to the time granularity Occurrence quantity distribution characteristics of being engaged in is vectorial, determines the business occurrence quantity in the predetermined amount of time.

In this specification embodiment, the history service data include first time nearest apart from the predetermined amount of time The the first history service data and the second history service data in addition to the first history service data of section,

The processing module 401, including：

In this specification embodiment, described device further includes：

In this specification embodiment, the similarity determining module determines described for passing through following any device Similarity between one feature vector and the second feature vector：The included angle cosine value of Euclidean distance, vector, and vector Absolute value of the difference.

In this specification embodiment, the business prediction of emergence size module 402, including：

Initial parameter optimizes unit, is used for the weight based on second feature vector described in the second feature vector sum, leads to It crosses loss function and predefined parameter optimization algorithm optimizes initial parameter, the initial parameter after being optimized；

In this specification embodiment, the predefined parameter optimization algorithm includes gradient descent algorithm, Newton method, quasi- newton Method, conjugate gradient method and heuristic value.

This specification embodiment provides a kind of prediction meanss of business occurrence quantity, by by the history before predetermined amount of time Business datum carries out sliding-model control, the business occurrence quantity vector of time granularity is obtained, furthermore it is also possible to according to history service The business occurrence quantity of continuous time in data generates business occurrence quantity distribution characteristics vector, finally, can be according to time granularity Business occurrence quantity vector sum business occurrence quantity distribution characteristics vector determine the business occurrence quantity in predetermined amount of time, in this way, logical Cross business occurrence quantity distribution characteristics vectorial (may include multidimensional characteristic vectors) come to the business occurrence quantity in predetermined amount of time into Row prediction, final prediction result can be accordingly adjusted in the case where short period service occurrence quantity fluctuates larger, effectively avoids industry Business occurrence quantity occurs in the case where short-term fluctuation larger impact final prediction result, moreover, being distributed by business occurrence quantity special The mode for levying vector, can also preferably capture variation tendency (such as history service number of business occurrence quantity in designated time period According to the variation tendency etc. of middle business occurrence quantity), so as to improve the predictablity rate of business occurrence quantity, reduce business risk with And improve fund utilization ratio.

Example IV

It is the prediction meanss for the business occurrence quantity that this specification embodiment provides above, is based on same thinking, this explanation Book embodiment also provides a kind of pre- measurement equipment of business occurrence quantity, as shown in Figure 5.

The pre- measurement equipment of the business occurrence quantity can be the server or terminal device that above-described embodiment provides.

The pre- measurement equipment of business occurrence quantity can generate bigger difference because configuration or performance are different, may include one Or more than one processor 501 and memory 502, one or more storage applications can be stored in memory 502 Program or data.Wherein, memory 502 can be of short duration storage or persistent storage.The application program for being stored in memory 502 can To include one or more modules (diagram is not shown), each module may include in the pre- measurement equipment to business occurrence quantity Series of computation machine executable instruction.Further, processor 501 could be provided as communicating with memory 502, in business The series of computation machine executable instruction in memory 502 is executed on the pre- measurement equipment of occurrence quantity.The prediction of business occurrence quantity is set Standby can also include one or more power supplys 503, one or more wired or wireless network interfaces 504, one or More than one input/output interface 505, one or more keyboards 506.

Specifically in the present embodiment, the pre- measurement equipment of business occurrence quantity include memory and one or more Program, either more than one program is stored in memory and one or more than one program may include for one of them One or more modules, and each module may include that series of computation machine in pre- measurement equipment to business occurrence quantity can It executes instruction, and is configured to by one that either more than one processor executes this or more than one program includes to be used for Carry out following computer executable instructions：

This specification embodiment provides a kind of pre- measurement equipment of business occurrence quantity, by by the history before predetermined amount of time Business datum carries out sliding-model control, the business occurrence quantity vector of time granularity is obtained, furthermore it is also possible to according to history service The business occurrence quantity of continuous time in data generates business occurrence quantity distribution characteristics vector, finally, can be according to time granularity Business occurrence quantity vector sum business occurrence quantity distribution characteristics vector determine the business occurrence quantity in predetermined amount of time, in this way, logical Cross business occurrence quantity distribution characteristics vectorial (may include multidimensional characteristic vectors) come to the business occurrence quantity in predetermined amount of time into Row prediction, final prediction result can be accordingly adjusted in the case where short period service occurrence quantity fluctuates larger, effectively avoids industry Business occurrence quantity occurs in the case where short-term fluctuation larger impact final prediction result, moreover, being distributed by business occurrence quantity special The mode for levying vector, can also preferably capture variation tendency (such as history service number of business occurrence quantity in designated time period According to the variation tendency etc. of middle business occurrence quantity), so as to improve the predictablity rate of business occurrence quantity, reduce business risk with And improve fund utilization ratio.

It is above-mentioned that this specification specific embodiment is described.Other embodiments are in the scope of the appended claims It is interior.In some cases, the action recorded in detail in the claims or step can be come according to different from the sequence in embodiment It executes and desired result still may be implemented.In addition, the process described in the accompanying drawings not necessarily require show it is specific suitable Sequence or consecutive order could realize desired result.In some embodiments, multitasking and parallel processing be also can With or it may be advantageous.

In the 1990s, the improvement of a technology can be distinguished clearly be on hardware improvement (for example, Improvement to circuit structures such as diode, transistor, switches) or software on improvement (improvement for method flow).So And with the development of technology, the improvement of current many method flows can be considered as directly improving for hardware circuit. Designer nearly all obtains corresponding hardware circuit by the way that improved method flow to be programmed into hardware circuit.Cause This, it cannot be said that the improvement of a method flow cannot be realized with hardware entities module.For example, programmable logic device (Programmable Logic Device, PLD) (such as field programmable gate array (Field Programmable Gate Array, FPGA)) it is exactly such a integrated circuit, logic function determines device programming by user.By designer Voluntarily programming comes a digital display circuit " integrated " on a piece of PLD, designs and makes without asking chip maker Dedicated IC chip.Moreover, nowadays, substitution manually makes IC chip, this programming is also used instead mostly " patrols Volume compiler (logic compiler) " software realizes that software compiler used is similar when it writes with program development, And the source code before compiling also write by handy specific programming language, this is referred to as hardware description language (Hardware Description Language, HDL), and HDL is also not only a kind of, but there are many kind, such as ABEL (Advanced Boolean Expression Language)、AHDL(Altera Hardware Description Language)、Confluence、CUPL(Cornell University Programming Language)、HDCal、JHDL (Java Hardware Description Language)、Lava、Lola、MyHDL、PALASM、RHDL(Ruby Hardware Description Language) etc., VHDL (Very-High-Speed are most generally used at present Integrated Circuit Hardware Description Language) and Verilog.Those skilled in the art also answer This understands, it is only necessary to method flow slightly programming in logic and is programmed into integrated circuit with above-mentioned several hardware description languages, The hardware circuit for realizing the logical method flow can be readily available.

Controller can be implemented in any suitable manner, for example, controller can take such as microprocessor or processing The computer for the computer readable program code (such as software or firmware) that device and storage can be executed by (micro-) processor can Read medium, logic gate, switch, application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), the form of programmable logic controller (PLC) and embedded microcontroller, the example of controller includes but not limited to following microcontroller Device：ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20 and Silicone Labs C8051F320, are deposited Memory controller is also implemented as a part for the control logic of memory.It is also known in the art that in addition to Pure computer readable program code mode is realized other than controller, can be made completely by the way that method and step is carried out programming in logic Controller is obtained in the form of logic gate, switch, application-specific integrated circuit, programmable logic controller (PLC) and embedded microcontroller etc. to come in fact Existing identical function.Therefore this controller is considered a kind of hardware component, and to including for realizing various in it The device of function can also be considered as the structure in hardware component.Or even, it can will be regarded for realizing the device of various functions For either the software module of implementation method can be the structure in hardware component again.

System, device, module or the unit that above-described embodiment illustrates can specifically realize by computer chip or entity, Or it is realized by the product with certain function.It is a kind of typically to realize that equipment is computer.Specifically, computer for example may be used Think personal computer, laptop computer, cellular phone, camera phone, smart phone, personal digital assistant, media play It is any in device, navigation equipment, electronic mail equipment, game console, tablet computer, wearable device or these equipment The combination of equipment.

For convenience of description, it is divided into various units when description apparatus above with function to describe respectively.Certainly, implementing this The function of each unit is realized can in the same or multiple software and or hardware when specification one or more embodiment.

It should be understood by those skilled in the art that, the embodiment of this specification can be provided as method, system or computer journey Sequence product.Therefore, complete hardware embodiment, complete software embodiment or knot can be used in this specification one or more embodiment The form of embodiment in terms of conjunction software and hardware.Moreover, this specification one or more embodiment can be used at one or more A wherein includes computer-usable storage medium (including but not limited to magnetic disk storage, the CD- of computer usable program code ROM, optical memory etc.) on the form of computer program product implemented.

The embodiment of this specification is with reference to the method, equipment (system) and computer journey according to this specification embodiment The flowchart and/or the block diagram of sequence product describes.It should be understood that flow chart and/or box can be realized by computer program instructions The combination of the flow and/or box in each flow and/or block and flowchart and/or the block diagram in figure.This can be provided A little computer program instructions are to all-purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices Processor to generate a machine so that pass through the finger that computer or the processor of other programmable data processing devices execute It enables and generates to specify in one flow of flow chart or multiple flows and/or one box of block diagram or multiple boxes The device of function.

These computer program instructions, which may also be stored in, can guide computer or other programmable data processing devices with spy Determine in the computer-readable memory that mode works so that instruction generation stored in the computer readable memory includes referring to Enable the manufacture of device, the command device realize in one flow of flow chart or multiple flows and/or one box of block diagram or The function of being specified in multiple boxes.

These computer program instructions also can be loaded onto a computer or other programmable data processing device so that count Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, in computer or The instruction executed on other programmable devices is provided for realizing in one flow of flow chart or multiple flows and/or block diagram one The step of function of being specified in a box or multiple boxes.

In a typical configuration, computing device includes one or more processors (CPU), input/output interface, net Network interface and memory.

Memory may include computer-readable medium in volatile memory, random access memory (RAM) and/or The forms such as Nonvolatile memory, such as read-only memory (ROM) or flash memory (flash RAM).Memory is computer-readable medium Example.

Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any method Or technology realizes information storage.Information can be computer-readable instruction, data structure, the module of program or other data. The example of the storage medium of computer includes, but are not limited to phase transition internal memory (PRAM), static RAM (SRAM), moves State random access memory (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), electric erasable Programmable read only memory (EEPROM), fast flash memory bank or other memory techniques, read-only disc read only memory (CD-ROM) (CD-ROM), Digital versatile disc (DVD) or other optical storages, magnetic tape cassette, tape magnetic disk storage or other magnetic storage apparatus Or any other non-transmission medium, it can be used for storage and can be accessed by a computing device information.As defined in this article, it calculates Machine readable medium does not include temporary computer readable media (transitory media), such as data-signal and carrier wave of modulation.

It should also be noted that, the terms "include", "comprise" or its any other variant are intended to nonexcludability Including so that process, method, commodity or equipment including a series of elements include not only those elements, but also wrap Include other elements that are not explicitly listed, or further include for this process, method, commodity or equipment intrinsic want Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that wanted including described There is also other identical elements in the process of element, method, commodity or equipment.

It will be understood by those skilled in the art that the embodiment of this specification can be provided as method, system or computer program production Product.Therefore, this specification one or more embodiment can be used complete hardware embodiment, complete software embodiment or combine software With the form of the embodiment of hardware aspect.Moreover, this specification one or more embodiment can be used it is one or more wherein The computer-usable storage medium for including computer usable program code (includes but not limited to magnetic disk storage, CD-ROM, light Learn memory etc.) on the form of computer program product implemented.

This specification one or more embodiment can computer executable instructions it is general on Described in hereafter, such as program module.Usually, program module includes executing particular task or realization particular abstract data type Routine, program, object, component, data structure etc..Can also put into practice in a distributed computing environment this specification one or Multiple embodiments, in these distributed computing environments, by being executed by the connected remote processing devices of communication network Task.In a distributed computing environment, the local and remote computer that program module can be located at including storage device is deposited In storage media.

Each embodiment in this specification is described in a progressive manner, identical similar portion between each embodiment Point just to refer each other, and each embodiment focuses on the differences from other embodiments.Especially for system reality For applying example, since it is substantially similar to the method embodiment, so description is fairly simple, related place is referring to embodiment of the method Part explanation.

The foregoing is merely the embodiments of this specification, are not intended to limit this application.For people in the art For member, the application can have various modifications and variations.It is all within spirit herein and principle made by it is any modification, etc. With replacement, improvement etc., should be included within the scope of claims hereof.

Claims

1. a kind of prediction technique of business occurrence quantity, the method includes：

History service data before predetermined amount of time are subjected to sliding-model control, obtain the business occurrence quantity of time granularity to Amount, and according to the business occurrence quantity of continuous time in the history service data, generate business occurrence quantity distribution characteristics vector；

Business occurrence quantity distribution characteristics vector, determines described pre- described in business occurrence quantity vector sum according to the time granularity The business occurrence quantity fixed time in section.

2. according to the method described in claim 1, the history service data include apart from the predetermined amount of time it is nearest the The first history service data of one period and the second history service data in addition to the first history service data,

The business of continuous time in history service data described in the business occurrence quantity vector sum according to the time granularity Occurrence quantity generates business occurrence quantity distribution characteristics vector, including：

According to the business occurrence quantity of continuous time in the first history service data, generate the distribution of business occurrence quantity first is special Sign vector；

According to the business occurrence quantity of continuous time in the second history service data, generate the distribution of business occurrence quantity second is special Sign vector.

3. according to the method described in claim 2, industry described in the business occurrence quantity vector sum according to the time granularity Occurrence quantity distribution characteristics of being engaged in is vectorial, and before determining the business occurrence quantity in the predetermined amount of time, the method further includes：

According to the similarity between the first eigenvector and the second feature vector, the second feature vector is determined Weight.

4. according to the method described in claim 3, between the determination first eigenvector and the second feature vector Similarity, including：

The similarity between the first eigenvector and the second feature vector is determined by following any method：It is European The included angle cosine value of distance, vector, and vectorial absolute value of the difference.

5. according to the method described in claim 3, industry described in the business occurrence quantity vector sum according to the time granularity Occurrence quantity distribution characteristics of being engaged in is vectorial, determines the business occurrence quantity in the predetermined amount of time, including：

Respectively by the business occurrence quantity vector of the time granularity and the first eigenvector and second feature vector It merges, the first eigenvector after being merged and second feature vector；

Based on the weight of second feature vector described in the second feature vector sum, calculated by loss function and predefined parameter optimization Method optimizes initial parameter, the initial parameter after being optimized；

According to after optimization initial parameter and the first eigenvector, determine the business occurrence quantity in the predetermined amount of time.

6. according to the method described in claim 5, the predefined parameter optimization algorithm includes gradient descent algorithm, Newton method, intends Newton method, conjugate gradient method and heuristic value.

7. a kind of prediction meanss of business occurrence quantity, described device include：

Processing module obtains time granularity for the history service data before predetermined amount of time to be carried out sliding-model control Business occurrence quantity vector, and according to the business occurrence quantity of continuous time in the history service data, generate business Measure distribution characteristics vector；

Business prediction of emergence size module, for business occurrence quantity described in the business occurrence quantity vector sum according to the time granularity Distribution characteristics vector, determines the business occurrence quantity in the predetermined amount of time.

8. device according to claim 7, the history service data include apart from the predetermined amount of time it is nearest the The first history service data of one period and the second history service data in addition to the first history service data,

The processing module, including：

First eigenvector generation unit is used for the business occurrence quantity according to continuous time in the first history service data, The first eigenvector of generation business occurrence quantity distribution；

Second feature vector generation unit is used for the business occurrence quantity according to continuous time in the second history service data, The second feature vector of generation business occurrence quantity distribution.

9. device according to claim 8, described device further include：

Similarity determining module, for determining the similarity between the first eigenvector and the second feature vector；

Weight determination module, for according to the similarity between the first eigenvector and the second feature vector, determining The weight of the second feature vector.

10. device according to claim 9, the similarity determining module, for being determined by following any device Similarity between the first eigenvector and the second feature vector：The included angle cosine value of Euclidean distance, vector, and The absolute value of the difference of vector.

11. device according to claim 9, the business prediction of emergence size module, including：

Combining unit, for respectively by the business occurrence quantity vector of the time granularity and the first eigenvector and described Second feature vector merges, the first eigenvector after being merged and second feature vector；

Initial parameter optimizes unit, for based on second feature vector, passing through loss function and predefined parameter optimization algorithm Initial parameter is optimized, the initial parameter after being optimized；

Business prediction of emergence size unit, for according to after optimization initial parameter and the first eigenvector, determine described pre- The business occurrence quantity fixed time in section.

12. according to the devices described in claim 11, the predefined parameter optimization algorithm include gradient descent algorithm, Newton method, Quasi-Newton method, conjugate gradient method and heuristic value.

13. a kind of pre- measurement equipment of business occurrence quantity, the pre- measurement equipment of the business occurrence quantity include：

Processor；And

It is arranged to the memory of storage computer executable instructions, the executable instruction makes the processing when executed Device：