CN105095646A - Data prediction method and electronic device - Google Patents
Data prediction method and electronic device Download PDFInfo
- Publication number
- CN105095646A CN105095646A CN201510368997.0A CN201510368997A CN105095646A CN 105095646 A CN105095646 A CN 105095646A CN 201510368997 A CN201510368997 A CN 201510368997A CN 105095646 A CN105095646 A CN 105095646A
- Authority
- CN
- China
- Prior art keywords
- data
- unit time
- period
- time period
- predicted value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Abstract
A data prediction method and an electronic device are disclosed. The method can predict data to be generated within a present time interval according to generated data in a plurality of history time intervals, and each time interval is divided into a plurality of corresponding unit time frame. The method comprises predicting a predicted value of data to be generated within a present unit frame of a present time interval and/or later each time frame according to generated data in corresponding time frames of history time intervals; and modifying the predicted value according to a rate of drift between data actually generated in at least one unit time frame before the present unit time frame of the present time interval and a predicted value of data of at least one unit time frame, to obtain a modified predicted value.
Description
Technical field
The disclosure relates to data mining, more specifically, relate to a kind of can the data predication method of predicted data trend and electronic equipment more accurately.
Background technology
Current internet application has been deep into the every aspect such as work, life, study of people.A large amount of tissues, the data of personnel in conscious collection internet, and intention uses the development of historical data to future to predict.But, for internet data or electronic commerce data, to predict the outcome usually and actual conditions exist larger gap, and can not adjust in real time deviation.
Summary of the invention
Object of the present disclosure be at least in part to provide a kind of can the data predication method of predicted data trend and electronic equipment more accurately.
According to an aspect of the present disclosure, provide a kind of data predication method, according to the data occurred in several historical time intervals, the data that will occur in prediction current time interval, wherein each time interval is divided into the unit time period of multiple correspondence, the method comprises: according to the data occurred in unit time period corresponding in historical time interval, the predicted value of the data that will occur in current one period and/or follow-up constituent parts period in prediction current time interval; And according to the data occurred actual at least one unit time period before the current one period in current time interval relative to the deviation ratio of the predicted value of data in this at least one unit time period, predicted value is revised, obtain revising predicted value.
According to another aspect of the present disclosure, provide a kind of electronic equipment, comprising: data collector, for image data; Memory storage, for storing gathered data; Data processing equipment, for according to the data that occur in several historical time intervals gathered, the data that will occur in prediction current time interval, wherein each time interval is divided into the unit time period of multiple correspondence, wherein, data processing equipment is configured to: according to the data occurred in unit time period corresponding in historical time interval, the predicted value of the data that will occur in current one period and/or follow-up constituent parts period in prediction current time interval; And according to the data occurred actual at least one unit time period before the current one period in current time interval relative to the deviation ratio of the predicted value of data in this at least one unit time period, predicted value is revised, obtain revising predicted value.
According to embodiment of the present disclosure, except utilizing except historical data predicts, also based on real time data, predicted value is revised immediately.So, can predicted data trend more exactly, and therefore suitable adjustment System resource.
Accompanying drawing explanation
By referring to the description of accompanying drawing to disclosure embodiment, above-mentioned and other objects of the present disclosure, feature and advantage will be more clear, in the accompanying drawings:
Fig. 1 shows the block diagram of the electronic equipment according to disclosure embodiment;
Fig. 2 shows the process flow diagram of the data predication method according to disclosure embodiment.
Embodiment
Below, with reference to the accompanying drawings embodiment of the present disclosure is described.But should be appreciated that, these describe just exemplary, and do not really want to limit the scope of the present disclosure.In addition, in the following description, the description to known features and technology is eliminated, to avoid unnecessarily obscuring concept of the present disclosure.
Term is only used to describe specific embodiment as used herein, and is not intended to limit the disclosure.Word used herein " one ", " one (kind) " and " being somebody's turn to do " etc. also should comprise the meaning of " multiple ", " multiple ", unless the context clearly indicates otherwise.In addition, term " comprises ", indicates " comprising " etc. the existence of described feature, step, operation and/or parts as used herein, but does not get rid of and exist or add other features one or more, step, operation or parts.
All terms (comprising technology and scientific terminology) have usual the understood implication of those skilled in the art as used herein, unless otherwise defined.It should be noted that term used herein should be interpreted as having the implication consistent with the context of this instructions, and should not explain in idealized or too mechanical mode.
Shown in the drawings of some block schemes and/or process flow diagram.Should be understood that some square frames in block scheme and/or process flow diagram or its combination can be realized by computer program instructions.These computer program instructions can be supplied to the processor of multi-purpose computer, special purpose computer or other programmable data treating apparatus, thus these instructions can create the device for realizing function/operation illustrated in these block schemes and/or process flow diagram when being performed by this processor.
Therefore, technology of the present disclosure can the form of hardware and/or software (comprising firmware, microcode etc.) realize.In addition, technology of the present disclosure can take the form of the computer program stored on the computer-readable medium of instruction, and this computer program can use for instruction execution system or combined command executive system.In context of the present disclosure, computer-readable medium can be can comprise, store, transmit, propagate or the arbitrary medium of transfer instruction.Such as, computer-readable medium can include but not limited to electricity, magnetic, optical, electrical magnetic, infrared or semiconductor system, device, device or propagation medium.The concrete example of computer-readable medium comprises: magnetic memory apparatus, as tape or hard disk (HDD); Light storage device, as CD (CD-ROM); Storer, as random access memory (RAM) or flash memory; And/or wire/wireless communication link.
Fig. 1 shows the block diagram of the electronic equipment according to disclosure embodiment.
As shown in Figure 1, data collector 101, memory storage 103 and data processing equipment 105 can be comprised according to the electronic equipment 100 of this embodiment.Electronic equipment 100 can be implemented as various forms of computing platform, as server, computing machine and mobile terminal etc.
Data collector 101 for image data, such as internet data or electronic commerce data etc.Data collector 101 can directly from data source as internet carrys out image data.Or data collector 101 can be Data Input Interface, the image data from special data acquisition equipment or data acquisition mechanism can be input in electronic equipment 100 via this input interface.The data gathered can comprise the data of the particular characteristic reacting a certain system.Such as, this index can be portfolio, and therefore can the load of reactive system.When system is e-commerce system, that can collect in " contract note amount ", " conclusion of the business number of packages ", " turnover ", " under only measure ", " place an order number of packages ", " order value ", " pageview ", " access times " etc. is one or more.According to these data, the portfolio of this e-commerce system can be inferred, and therefore judge the load of this system, thus can according to load correspondingly adjustment System resource.
The data that memory storage 103 gathers for storing data collector 101.In addition, (such as, various program, as the program for realizing following methods for the information that the operation of all right store electrons equipment 100 of memory storage 103 is relevant; And the required or various data that produce in working procedure process).Memory storage 103 can be implemented as various volatibility and/or nonvolatile storage technologies, and memory storage can be comprised as hard disk, storage card etc., storer is as static RAM (SRAM), dynamic RAM (DRAM), flash memory etc.
Data processing equipment 105, for according to the data occurred in several historical time intervals gathered, predicts the data that will occur in current time interval.Data processing equipment 105 can be implemented as processor or microprocessor, such as CPU (central processing unit) (CPU).Predicting the outcome of data processing equipment 105 can output to electronic device exterior, such as, output to display to show.
Fig. 2 shows the process flow diagram of the data predication method according to disclosure embodiment.This data predication method 200 such as can be performed by data processing equipment 105.
As shown in Figure 2, the method is included in operation 201 and uses historical data to predict.According to embodiment of the present disclosure, time shaft is divided into " time interval ", that is, one specific period.Such as, this time interval can be 1 day, 1 week, 1 ten days, January, the first quarter, 1 year etc.For current time interval (such as, the same day, this week, this ten days, this month, current season, this year etc.), time interval (such as, yesterday, last week, the first tenday period of a month, last month, preceding quarter, last year etc.) before all becomes " history ".So, can according to historical data, the data particularly occurred in several historical time intervals, predict the data that will occur in current time interval.In order to embody the recent trend of system better, these several contiguous current times in historical time interval are interval.In addition, these several historical time intervals can be continuous print in time, also can be to be separated.Preferably, these several historical time intervals can be immediately preceding front interval multiple continuous times, current time interval (such as, the same day) (such as, from the 7th day before the same day until continuous 7 days altogether yesterday).
In addition, each time interval can also be divided into multiple unit time period.Such as, when time interval is 1 day, can by the hour or dividing unit period half an hour (now, each time interval is divided into 24 or 48 unit time period).The granularity of division of this time interval and unit time period can be specified according to the characteristic of such as involved system.
At this, can according to the data occurred in unit time period corresponding in historical time interval, unit section and/or the predicted value of data that will occur in the follow-up constituent parts period time current in prediction current time interval.Various ways can be had to carry out this prediction.Such as, for the constituent parts period in current time interval, based on the weighted mean value of the data occurred in unit time period corresponding in historical time interval, the predicted value of the data that will occur in this unit time period can be predicted.
Such as, suppose to there is (M-1) individual historical time interval and interval M the time interval altogether of current time, and each time interval is divided into N number of unit time period, if the data occurred in a jth unit time period in i-th time interval are ADd
it
j, comparable data is CDd
it
j.Here, time interval and unit time period arrange all in chronological order, thus " i-th time interval " and " a jth unit time period " has the implication determined.In the following way, weighted mean value can be obtained:
1S＜M，1≤j≤N
CDd
it
j=XAdd
(i-1)t
j+ YCDd
(i-1)t
j, 2≤i≤M, 1≤j≤N wherein, CDd
mt
jas the predicted value of data in a jth unit time period in current time interval, α
1, α
2..., α
sand X, Y are weighting coefficient, and α
1+ α
2+ ... + α
s=S, and X+Y=1.These weighted datas can be obtained by regretional analysis according to historical data.
After acquisition predicted value, the method proceeds to operation 203, according to real time data, revises predicted value.At this, so-called " in real time " data, refer to that the data that occur in current time interval (such as, the data that the same day occurs,, because not yet there are any data in present period to be predicted and subsequent time intervals in the data occurred in the unit time period particularly before the current one period).Usually, in conventional Forecasting Methodology, only consider " history " data, and do not consider this " in real time " data.
According to embodiment of the present disclosure, can according to the deviation ratio of the data occurred actual at least one unit time period before the current one period in current time interval relative to the predicted value of data in this at least one unit time period, predicted value is revised, obtains revising predicted value.At this, for the constituent parts period in current time interval, identical Forecasting Methodology can be utilized to obtain its predicted value according to historical data.Like this, described deviation ratio reacted this Forecasting Methodology be applied to current time interval time extent of deviation.Therefore, the predicted value of this deviation ratio to current one period and follow-up unit time period can be utilized to revise.Preferably, the deviation ratio in all unit time period before the current one period in current time interval can be used for revising.In one example, the deviation ratio in a unit time period is defined as the actual data of generation and the ratio of the predicted value of data in this unit time period in this unit time period.Such as, for a jth unit time period in current time interval, deviation ratio can be expressed as 1/R
j, wherein R
j=CDd
mt
j÷ Add
mt
j, 1≤j≤N.This fixed case is as undertaken by the mean value of at least one unit time period deviation ratio separately as described in predicted value being multiplied by.
In addition, deviation ratio in close to one or more unit time period (particularly immediately preceding the first period before the current one period and/or the second period) of current one period in the constituent parts period indicate the data of actual generation this unit time period in and for this unit time period predicted value between skew do not exceed predetermined threshold time (also, show that the precision of prediction of the Forecasting Methodology of current use is higher), can be used as to revise predicted value by approximate for predicted value.
Such as, for current one period P or follow-up unit time period in current time interval, can revise as follows:
If | 1-R
p-1|≤0.1 or | 1-R
p-2|≤0.1, then FD
j=CDd
mt
j, P≤j≤N; If | 1-R
p-1| > 0.1 and | 1-R
p-2| > 0.1, then
wherein, FD
jrepresent the correction predicted value of jth unit time period.
In this example, with " 0.1 " for threshold value judges that whether deviation ratio is enough little, namely whether Forecasting Methodology is enough accurate.But the disclosure is not limited thereto.According to practical situations, this threshold value can be determined adaptively.
Below, by concrete for description one example, to understand above-mentioned data predication method better.In this example, suppose time interval for " my god ", and use the historical data in first 7 days of the same day and the data on the same day (that is, M=8), unit time period is " half an hour " (being divided into 48 unit time period for one day, that is, N=48).
First 7 days as shown in table 1 below with the real data on the same day.These data are such as collected by the data collector 101 shown in Fig. 1.
Table 1
| Time period | 1st day | 2nd day | 3rd day | 4th day | 5th day | 6th day | 7th day | The same day |
| 00：00～00：30 | ADd 1t 1 | ADd 2t 1 | ADd 3t 1 | ADd 4t 1 | ADd 5t 1 | ADd 6t 1 | ADd 7t 1 | ADd 8t 1 |
| 00：30～01：00 | ADd 1t 2 | ADd 2t 2 | ADd 3t 2 | ADd 4t 2 | ADd 5t 2 | ADd 6t 2 | ADd 7t 2 | ADd 8t 2 |
| 01：00～01：30 | ADd 1t 3 | ADd 2t 3 | ADd 3t 3 | ADd 4t 3 | ADd 5t 3 | ADd 6t 3 | ADd 7t 3 | ADd 8t 3 |
| 01：30～02：00 | ADd 1t 4 | ADd 2t 4 | ADd 3t 4 | ADd 4t 4 | ADd 5t 4 | ADd 6t 4 | ADd 7t 4 | ADd 8t 4 |
| 02：00～02：30 | ADd 1t 5 | ADd 2t 5 | ADd 3t 5 | ADd 4t 5 | ADd 5t 5 | ADd 6t 5 | ADd 7t 5 | ADd 8t 5 |
| …… | …… | …… | …… | …… | …… | …… | …… | …… |
| 23：30～00：00 | ADd 1t 48 | ADd 2t 48 | ADd 3t 48 | ADd 4t 48 | ADd 5t 48 | ADd 6t 48 | ADd 7t 48 | ADd 8t 48 |
Shown in the comparable data following table 2 on first 7 days and the same day.
Table 2
| Time period | 1st day | 2nd day | 3rd day | 4th day | 5th day | 6th day | 7th day | The same day |
| 00：00～00：30 | CDd 1t 1 | CDd 2t 1 | CDd 3t 1 | CDd 4t 1 | CDd 5t 1 | CDd 6t 1 | CDd 7t 1 | CDd 8t 1 |
| 00：30～01：00 | CDd 1t 2 | CDd 2t 2 | CDd 3t2 | CDd 4t 2 | CDd 5t 2 | CDd 6t 2 | CDd 7t 2 | CDd 8t 2 |
| 01：00～01：30 | CDd 1t 3 | CDd 2t 3 | CDd 3t 3 | CDd 4t 3 | CDd 5t 3 | CDd 6t 3 | CDd 7t 3 | CDd 8t 3 |
| 01：30～02：00 | CDd 1t 4 | CDd 2t 4 | CDd 3t 4 | CDd 4t 4 | CDd 5t 4 | CDd 6t 4 | CDd 7t 4 | CDd 8t 4 |
| 02：00～02：30 | CDd 1t 5 | CDd 2t 5 | CDd 3t 5 | CDd 4t 5 | CDd 5t 5 | CDd 6t 5 | CDd 7t 5 | CDd 8t 5 |
| …… | …… | …… | …… | …… | …… | …… | …… | …… |
| 23：30～00：00 | CDd 1t 48 | CDd 2t 48 | CDd 3t 48 | CDd 4t 48 | CDd 5t 48 | CDd 6t 48 | CDd 7t 48 | CDd 8t 48 |
In this example, comparable data calculates as follows:
CDd
it
j＝X·ADd
i-1t
j+Y·CDd
i-1t
j，2≤i≤8，1≤j≤48。Wherein, α, β, γ, X, Y are weighting coefficient, alpha+beta+γ=3, X+Y=1.If do not have a large amount of historical data that regretional analysis can not be carried out, be then initialized as α=1, β=1, γ=1, X=0.4, Y=0.6; If there has been a large amount of historical data to carry out regretional analysis, then α, β, γ, X, Y value have adopted regretional analysis, find out comparable data and real data closest to the numerical value in situation.
In this example, for the constituent parts period in the 1st day, comparable data CDd
1t
jcarry out calculating (that is, S=3) according to the data in first three day.But the disclosure is not limited thereto, the value of S differently can be set according to actual conditions.
According to above calculating, the comparable data CDd of day part on the same day can be obtained
8t
j, as (roughly) predicted value of day part.
Next, can revise (roughly) predicted value obtained according to real time data.
Particularly, for the 1st unit time period on the same day, when this period starts, i.e. 00:00 point, the data that will occur in prediction (that is, (roughly) predicted value being revised further, to obtain more accurate predicted value) this unit time period.Now, just started and there is not any real data the same day.Therefore, can by predicted value CDd
8t
1be used as to revise predicted value FD
1.
For the 2nd unit time period on the same day, can when this period start, namely 00:30 point, predicts the data that will occur in this unit time period.At this, deviation ratio in unit time period on the same day the 1st can be utilized as mentioned above to predicted value CDd
8t
2revise.Alternatively, owing to now only having carried out single unit time period, so deviation ratio statistically may not have meaning.Therefore, also can directly by predicted value CDd
8t
2be used as to revise predicted value FD
2.
For other follow-up unit time period j (3≤j≤48) on the same day, when this period starts, predict the data that will occur in this unit time period (now, this period is current one period P) and/or its follow-up unit time period.As mentioned above, this can carry out as follows:
If | 1-R
p-1|≤0.1 or | 1-R
p-2|≤0.1, then FD
j=CDd
8t
j, P≤j≤48;
If | 1-R
p-1| > 0.1 and | 1-R
p-2| > 0.1, then
In this example, at current one period P, also predict and have modified the data in its follow-up constituent parts period.Correction predicted value in these follow-up unit time period can for providing Data support with the system decision-making.Alternatively, at current one period P, can not carry out predicting and revising to the follow-up unit time period of part (such as, every a unit time period) to its follow-up unit time period or only.These follow-up unit time period can carry out predicting and revising when the time proceeds to this unit time period (that is, this unit time period self becomes present period) again.
Technology disclosed herein can have multiple application, such as, can be applied to e-commerce system.In regular situation, sale that the same day, whole day can reach, flow value can be learnt according to the real data in early stage, sales forecast is carried out to ecommerce operation personnel, understand business development in time and be very helpful.In the great advertising campaign of special circumstances as ecommerce, real time data fluctuation is very large, on the basis of predicted data, has real time data according to the same day, and immediately revises predicted data.Like this, can suitable mixing system resource, such as can open standby server to avoid traffic congestion when predicting that portfolio is larger, and server can be changed into low-power mode or closed portion server to reduce power consumption when predicting that portfolio is less, thus the optimization use of system resource can be realized.And, operation personnel can be made to have good precognition for the effect of advertising campaign, contribute to the adjustment of promotion plan, to reach good animation effect, realize auxiliary operation, decision-making management.
Above embodiment of the present disclosure is described.But these embodiments are only used to the object illustrated, and are not intended to limit the scope of the present disclosure.Although respectively describing each embodiment above, this is not also meaning that the measure in each embodiment can not advantageously be combined.The scope of the present disclosure is by claims and equivalents thereof.Do not depart from the scope of the present disclosure, those skilled in the art can make multiple substituting and amendment, and these substitute and amendment all should fall within the scope of the present disclosure.
Claims (15)
1. a data predication method, according to the data occurred in several historical time intervals, the data that will occur in prediction current time interval, wherein each time interval is divided into the unit time period of multiple correspondence, and the method comprises:
According to the data occurred in unit time period corresponding in historical time interval, the predicted value of the data that will occur in current one period and/or follow-up constituent parts period in prediction current time interval; And
According to the data occurred actual at least one unit time period before the current one period in current time interval relative to the deviation ratio of the predicted value of data in this at least one unit time period, predicted value is revised, obtain revising predicted value.
2. data predication method according to claim 1, wherein, the deviation ratio of constituent parts period is represented with the ratio of the predicted value of data in this unit time period by the data occurred actual in this unit time period.
3. data predication method according to claim 2, wherein, by predicted value being multiplied by the mean value of described at least one unit time period deviation ratio separately, obtains revising predicted value.
4. data predication method according to claim 1, wherein, deviation ratio in close to one or more unit time period of current one period in the constituent parts period indicate the actual data that occur in this unit time period and for this unit time period predicted value between skew do not exceed predetermined threshold time, usage forecastings value is as correction predicted value.
5. data predication method according to claim 1, wherein, for the constituent parts period in current time interval, based on the weighted mean value of the data occurred in unit time period corresponding in historical time interval, predict the predicted value of the data that will occur in this unit time period.
6. data predication method according to claim 5, wherein, there is (M-1) individual historical time interval and interval M the time interval altogether of current time, and each time interval is divided into N number of unit time period, if the data occurred in a jth unit time period in i-th time interval are ADd
it
j, comparable data is CDd
it
j, then in the following way, obtain weighted mean value:
CDd
it
j＝X·Add
(i-1)t
j+Y·CDd
(i-1)t
j，2≤i≤M，1≤j≤N
Wherein, CDd
mt
jas the predicted value of data in a jth unit time period in current time interval, α
1, α
2..., α
sand X, Y are weighting coefficient, and α
1+ α
2+ ... + α
s=S, and X+Y=1.
7. data predication method according to claim 6, wherein, according to historical data by regretional analysis, obtains weighting coefficient.
8. data predication method according to claim 6, wherein, for a jth unit time period in current time interval, deviation ratio is expressed as 1/R
j, wherein:
R
j＝CDd
Mt
j÷Add
Mt
j，1≤j≤N。
9. data predication method according to claim 8, wherein, for current one period P or follow-up unit time period in current time interval, revise as follows:
If | 1-R
p-1|≤0.1 or | 1-R
p-2|≤0.1, then FD
j=CDd
mt
j, P≤j≤N; If | 1-R
p-1| > 0.1 and | 1-R
p-2| > 0.1, then
p≤j≤N, wherein, FD
jrepresent the correction predicted value of jth unit time period.
10. data predication method according to claim 1, also comprises: according to correction predicted value, allocating system resource.
11. 1 kinds of electronic equipments, comprising:
Data collector, for image data;
Memory storage, for storing gathered data;
Data processing equipment, for according to the data that occur in several historical time intervals gathered, the data that will occur in prediction current time interval, wherein each time interval is divided into the unit time period of multiple correspondence,
Wherein, data processing equipment is configured to:
According to the data occurred in unit time period corresponding in historical time interval, the predicted value of the data that will occur in current one period and/or follow-up constituent parts period in prediction current time interval; And
According to the data occurred actual at least one unit time period before the current one period in current time interval relative to the deviation ratio of the predicted value of data in this at least one unit time period, predicted value is revised, obtain revising predicted value.
12. electronic equipments according to claim 11, wherein, the deviation ratio of constituent parts period is represented with the ratio of the predicted value of data in this unit time period by the data occurred actual in this unit time period.
13. electronic equipments according to claim 12, wherein, data processing equipment, by predicted value being multiplied by the mean value of described at least one unit time period deviation ratio separately, obtains revising predicted value.
14. electronic equipments according to claim 11, wherein, deviation ratio in close to one or more unit time period of current one period in the constituent parts period indicate the actual data that occur in this unit time period and for this unit time period predicted value between skew do not exceed predetermined threshold time, data processing equipment usage forecastings value is as correction predicted value.
15. electronic equipments according to claim 11, wherein, for the constituent parts period in current time interval, data processing equipment, based on the weighted mean value of the data occurred in unit time period corresponding in historical time interval, predicts the predicted value of the data that will occur in this unit time period.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510368997.0A CN105095646A (en) | 2015-06-29 | 2015-06-29 | Data prediction method and electronic device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510368997.0A CN105095646A (en) | 2015-06-29 | 2015-06-29 | Data prediction method and electronic device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105095646A true CN105095646A (en) | 2015-11-25 |
Family
ID=54576064
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510368997.0A Pending CN105095646A (en) | 2015-06-29 | 2015-06-29 | Data prediction method and electronic device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105095646A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107507030A (en) * | 2017-08-22 | 2017-12-22 | 北京京东尚科信息技术有限公司 | The method and apparatus of information prediction |
| CN108596652A (en) * | 2018-03-28 | 2018-09-28 | 麒麟合盛网络技术股份有限公司 | Active users prediction technique and device |
| CN109598620A (en) * | 2018-12-03 | 2019-04-09 | 南方电网能源发展研究院有限责任公司 | Estimation method, device, computer equipment and the storage medium of investment project |
| CN110046151A (en) * | 2019-03-05 | 2019-07-23 | 努比亚技术有限公司 | A kind of data cleaning method, server and computer readable storage medium |
| CN110858912A (en) * | 2018-08-22 | 2020-03-03 | 中国电信股份有限公司 | Streaming media caching method and system, caching policy server and streaming service node |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6430539B1 (en) * | 1999-05-06 | 2002-08-06 | Hnc Software | Predictive modeling of consumer financial behavior |
| CN101916335A (en) * | 2010-08-19 | 2010-12-15 | 河北农业大学 | Prediction method of city water-requirement time series-exponent smoothing model |
| CN102968670A (en) * | 2012-10-23 | 2013-03-13 | 北京京东世纪贸易有限公司 | Method and device for predicting data |
| CN103886018A (en) * | 2014-02-21 | 2014-06-25 | 车智互联（北京）科技有限公司 | Data predication device, data predication method and electronic equipment |
| CN104182801A (en) * | 2013-05-22 | 2014-12-03 | 阿里巴巴集团控股有限公司 | Method and device for predicting website visits |
-
2015
- 2015-06-29 CN CN201510368997.0A patent/CN105095646A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6430539B1 (en) * | 1999-05-06 | 2002-08-06 | Hnc Software | Predictive modeling of consumer financial behavior |
| CN101916335A (en) * | 2010-08-19 | 2010-12-15 | 河北农业大学 | Prediction method of city water-requirement time series-exponent smoothing model |
| CN102968670A (en) * | 2012-10-23 | 2013-03-13 | 北京京东世纪贸易有限公司 | Method and device for predicting data |
| CN104182801A (en) * | 2013-05-22 | 2014-12-03 | 阿里巴巴集团控股有限公司 | Method and device for predicting website visits |
| CN103886018A (en) * | 2014-02-21 | 2014-06-25 | 车智互联（北京）科技有限公司 | Data predication device, data predication method and electronic equipment |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107507030A (en) * | 2017-08-22 | 2017-12-22 | 北京京东尚科信息技术有限公司 | The method and apparatus of information prediction |
| CN108596652A (en) * | 2018-03-28 | 2018-09-28 | 麒麟合盛网络技术股份有限公司 | Active users prediction technique and device |
| CN110858912A (en) * | 2018-08-22 | 2020-03-03 | 中国电信股份有限公司 | Streaming media caching method and system, caching policy server and streaming service node |
| CN109598620A (en) * | 2018-12-03 | 2019-04-09 | 南方电网能源发展研究院有限责任公司 | Estimation method, device, computer equipment and the storage medium of investment project |
| CN110046151A (en) * | 2019-03-05 | 2019-07-23 | 努比亚技术有限公司 | A kind of data cleaning method, server and computer readable storage medium |
| CN110046151B (en) * | 2019-03-05 | 2023-08-11 | 努比亚技术有限公司 | Data cleaning method, server and computer readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chien et al. | Strategic capacity planning for smart production: Decision modeling under demand uncertainty | |
| Wu et al. | Fractional ARIMA with an improved cuckoo search optimization for the efficient Short-term power load forecasting | |
| Ziarnetzky et al. | Rolling horizon, multi-product production planning with chance constraints and forecast evolution for wafer fabs | |
| Khandelwal et al. | Amazon EC2 spot price prediction using regression random forests | |
| CN105095646A (en) | Data prediction method and electronic device | |
| US9753782B2 (en) | Resource consumption optimization | |
| US20130110756A1 (en) | Short-term Load Forecast Using Support Vector Regression and Feature Learning | |
| US20140257778A1 (en) | Devices for Forecasting Ratios in Hierarchies | |
| JP2013137759A (en) | System and method for generating marketing mix solution | |
| Yan et al. | Mid-term electricity market clearing price forecasting utilizing hybrid support vector machine and auto-regressive moving average with external input | |
| Sawik | A stochastic optimisation approach to maintain supply chain viability under the ripple effect | |
| Barnett et al. | Nonlinear and complex dynamics in economics | |
| US20130332243A1 (en) | Predictive analytics based ranking of projects | |
| US9449049B2 (en) | Returning estimated value of search keywords of entire account | |
| US20090112691A1 (en) | System and method for scheduling online keyword auctions over multiple time periods subject to budget and query volume constraints | |
| AU2018217286A1 (en) | Control system with machine learning time-series modeling | |
| US20150294350A1 (en) | Automated optimization of a mass policy collectively performed for objects in two or more states and a direct policy performed in each state | |
| US10963799B1 (en) | Predictive data analysis of stocks | |
| US20140289007A1 (en) | Scenario based customer lifetime value determination | |
| JP2018180996A (en) | Prediction device, prediction method, and prediction program | |
| US20180329374A1 (en) | Systems and methods for optimizing an energy storage device charging schedule | |
| US20160140577A1 (en) | Unified marketing model based on conduit variables | |
| US20150186928A1 (en) | Real-time marketing portfolio optimization and reapportioning | |
| Kapindula et al. | The effects of external debt servicing on infrastructure spending: A case of Zambia | |
| Aggarwal et al. | Day-ahead price forecasting in Ontario electricity market using variable-segmented support vector machine-based model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20151125 |
|
| RJ01 | Rejection of invention patent application after publication |