CN111339156A - Long-term determination method and device of business data and computer readable storage medium - Google Patents
Long-term determination method and device of business data and computer readable storage medium Download PDFInfo
- Publication number
- CN111339156A CN111339156A CN202010091464.3A CN202010091464A CN111339156A CN 111339156 A CN111339156 A CN 111339156A CN 202010091464 A CN202010091464 A CN 202010091464A CN 111339156 A CN111339156 A CN 111339156A
- Authority
- CN
- China
- Prior art keywords
- time period
- value
- service data
- time
- business data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2462—Approximate or statistical queries
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Probability & Statistics with Applications (AREA)
- Fuzzy Systems (AREA)
- Computational Linguistics (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The disclosure relates to a long-term determination method and device of business data and a computer readable storage medium, and relates to the technical field of computers. The method of the present disclosure comprises: dividing each first time period corresponding to the service data to be determined into one or more groups, wherein each group corresponds to a second time period; the first time period and the second time period are divided according to time units with different granularities respectively; for each second time period, determining the service data statistic value in each second time period in a plurality of second time periods before the second time period according to the service data statistic value in the second time period; and for each first time period, determining the service data value of the first time period according to the service data statistic value in the second time period corresponding to the first time period, thereby obtaining the value of the service data to be determined of each first time period.
Description
Technical Field
The present disclosure relates to the field of computer technologies, and in particular, to a method and an apparatus for determining service data for a long time, and a computer-readable storage medium.
Background
With the development of computer technology, the data volume has been explosively increased. Analysis of the large amount of data can be used to guide future production activities.
For the service data, a time sequence model is generally adopted to determine the service data in the short term in the future.
Disclosure of Invention
The inventor finds that the current method for determining the business data is mainly used for determining the short-term fine-grained time, and is inaccurate if the future long-term business data is determined.
One technical problem to be solved by the present disclosure is: the accuracy of long-term service data determination is improved.
According to some embodiments of the present disclosure, a method for determining long-term service data is provided, including: dividing each first time period corresponding to the service data to be determined into one or more groups, wherein each group corresponds to a second time period; the first time period and the second time period are divided according to time units with different granularities respectively; for each second time period, determining the service data statistic value in each second time period in a plurality of second time periods before the second time period according to the service data statistic value in the second time period; and for each first time period, determining the service data value of the first time period according to the service data statistic value in the second time period corresponding to the first time period, thereby obtaining the value of the service data to be determined of each first time period.
In some embodiments, for each second time period, determining the traffic data statistics value in each second time period in a plurality of second time periods before the second time period comprises: determining a plurality of second time periods within a preset time window length range before the second time period and a service data statistic value in each second time period; inputting the service data statistic value in each second time period into a pre-trained coarse-grained determination model to obtain an output service data statistic value in the second time period; wherein, the coarse-grained determination model comprises: a ridge regression model, a timing model, or a random forest model.
In some embodiments, for each first time period, determining the traffic data value of the first time period according to the traffic data statistic value in the second time period corresponding to the first time period includes: according to the time sequence of the first time period in the corresponding second time period, determining a pre-trained fine-grained determination model corresponding to the first time period; inputting the business data statistic value in a second time period corresponding to the first time period into a fine-grained determination model corresponding to the first time period to obtain an output business data value of the first time period; the fine-grained determination models corresponding to the first time periods in different time sequences have different parameters, and the fine-grained determination models are used for expressing the relation between the business data values of the first time periods and the business data statistical values in the corresponding second time periods.
In some embodiments, for each first time period, determining the traffic data value of the first time period according to the traffic data statistic value in the second time period corresponding to the first time period includes: adjusting the service data statistics value in the second time period corresponding to the first time period according to the service data value in each first time period before the first time period belonging to the same second time period and the service data statistics value in the second time period corresponding to the first time period; wherein the traffic data value of each first time period before the first time period is determined; and determining the business data value of the first time period according to the adjusted business data statistic value in the second time period corresponding to the first time period.
In some embodiments, the service data statistics in the second time period corresponding to the first time period are adjusted according to the following formula:
wherein, P represents the business data statistic value in the second time period corresponding to the first time period, and PmThe value of the traffic data of the mth first time period before the first time period belonging to the same second time period is represented, N represents the number of the first time periods included in one second time period, and M represents the number of the first time periods before the first time period belonging to the same second time period.
In some embodiments, determining the service data value of the first time period according to the adjusted service data statistic value in the second time period corresponding to the first time period includes: according to the time sequence of the first time period in the corresponding second time period, determining a pre-trained fine-grained determination model corresponding to the first time period; inputting the adjusted business data statistic value in a second time period corresponding to the first time period into a fine-grained determination model corresponding to the first time period to obtain an output business data value of the first time period; the fine-grained determination models corresponding to the first time periods in different time sequences have different parameters, and the fine-grained determination models are used for expressing the relation between the business data values of the first time periods and the business data statistical values in the corresponding second time periods.
In some embodiments, the method further comprises: obtaining historical business data values of all first time periods and business data statistics values of all second time periods; training the coarse-grained determination model according to the statistical value of the business data in each historical second time period; the coarse granularity determination model is used for representing the relation between the business data statistic value in one second time period and the business data statistic value in each second time period in a plurality of second time periods before the second time period; and training a fine-grained determination model according to the historical business data values of the first time periods and the historical business data statistical values of the second time periods, wherein the fine-grained determination model is used for expressing the relation between the business data values of the first time periods and the corresponding business data statistical values in the second time periods.
In some embodiments, the method further comprises: determining the average value and the variance of the business data values of each historical first time period; determining whether the service data value of each first time period is abnormal or not according to the average value and the variance aiming at the service data value of each first time period; and under the condition that the service data value of the first time period is abnormal, deleting the service data value of the first time period, and taking the service data value of the first time period before the first time period as the service data value of the first time period.
In some embodiments, the traffic data includes: price of the commodity, sales volume of the commodity, visit volume of the website, or traffic volume of the website.
In some embodiments, the method further comprises: under the condition that the business data is the price of the commodity, the sales volume of the commodity or the access volume of the commodity, recommending the commodity to the user according to the value of the business data to be determined in each first time period; or, under the condition that the service data is the access volume of the website or the flow of the website, adjusting the resource deployment corresponding to the website according to the value of the service data to be determined in each first time period.
According to other embodiments of the present disclosure, there is provided a long-term service data determining apparatus, including: the time division module is used for dividing each first time period corresponding to the service data to be determined into one or more groups, and each group corresponds to a second time period; the first time period and the second time period are divided according to time units with different granularities respectively; the coarse granularity determining module is used for determining a service data statistic value in each second time period in a plurality of second time periods before the second time period according to the service data statistic value in each second time period; and the fine granularity determining module is used for determining the business data value of each first time period according to the business data statistic value in the second time period corresponding to the first time period so as to obtain the value of the business data to be determined of each first time period.
In some embodiments, the coarse granularity determination module is configured to determine a plurality of second time periods within a preset time window length range before the second time period, and a traffic data statistic value in each second time period; inputting the service data statistic value in each second time period into a pre-trained coarse-grained determination model to obtain an output service data statistic value in the second time period; wherein, the coarse-grained determination model comprises: a ridge regression model, a timing model, or a random forest model.
In some embodiments, the fine-grained determination module is configured to determine a pre-trained fine-grained determination model corresponding to the first time period according to a time sequence of the first time period in a corresponding second time period; inputting the business data statistic value in a second time period corresponding to the first time period into a fine-grained determination model corresponding to the first time period to obtain an output business data value of the first time period; the fine-grained determination models corresponding to the first time periods in different time sequences have different parameters, and the fine-grained determination models are used for expressing the relation between the business data values of the first time periods and the business data statistical values in the corresponding second time periods.
In some embodiments, the fine-grained determination module is configured to adjust the business data statistics value in the second time period corresponding to the first time period according to the business data value in each first time period before the first time period belonging to the same second time period and the business data statistics value in the second time period corresponding to the first time period; wherein the traffic data value of each first time period before the first time period is determined; and determining the business data value of the first time period according to the adjusted business data statistic value in the second time period corresponding to the first time period.
In some embodiments, the service data statistics in the second time period corresponding to the first time period are adjusted according to the following formula:
wherein, P represents the business data statistic value in the second time period corresponding to the first time period, and PmRepresenting a traffic data value of an M-th first time period before the first time period belonging to the same second time period, N representing the number of first time periods included in one second time period, M representing the number of first time periods included in one second time periodThe number of first time periods preceding the first time period belonging to the same second time period.
In some embodiments, the fine-grained determination module is configured to determine a pre-trained fine-grained determination model corresponding to the first time period according to a time sequence of the first time period in a corresponding second time period; inputting the adjusted business data statistic value in a second time period corresponding to the first time period into a fine-grained determination model corresponding to the first time period to obtain an output business data value of the first time period; the fine-grained determination models corresponding to the first time periods in different time sequences have different parameters, and the fine-grained determination models are used for expressing the relation between the business data values of the first time periods and the business data statistical values in the corresponding second time periods.
In some embodiments, the apparatus further comprises: the training module is used for acquiring historical business data values of all the first time periods and business data statistical values of all the second time periods; training the coarse-grained determination model according to the statistical value of the business data in each historical second time period; the coarse granularity determination model is used for representing the relation between the business data statistic value in one second time period and the business data statistic value in each second time period in a plurality of second time periods before the second time period; and training a fine-grained determination model according to the historical business data values of the first time periods and the historical business data statistical values of the second time periods, wherein the fine-grained determination model is used for expressing the relation between the business data values of the first time periods and the corresponding business data statistical values in the second time periods.
In some embodiments, the apparatus further comprises: the data processing module is used for determining the average value and the variance of the business data values of each historical first time period; determining whether the service data value of each first time period is abnormal or not according to the average value and the variance aiming at the service data value of each first time period; and under the condition that the service data value of the first time period is abnormal, deleting the service data value of the first time period, and taking the service data value of the first time period before the first time period as the service data value of the first time period.
In some embodiments, the traffic data includes: price of the commodity, sales volume of the commodity, visit volume of the website, or traffic volume of the website.
In some embodiments, the apparatus further comprises: the recommending module is used for recommending commodities to the user according to the value of the to-be-determined business data of each first time period under the condition that the business data are the price of the commodities, the sales volume of the commodities or the access volume of the commodities; or, the resource scheduling module is configured to, when the service data is the access volume of the website or the traffic volume of the website, adjust resource deployment corresponding to the website according to the value of the service data to be determined in each first time period.
According to still other embodiments of the present disclosure, there is provided a long-term determining apparatus for service data, including: a processor; and a memory coupled to the processor for storing instructions that, when executed by the processor, cause the processor to perform a method of long-term determination of traffic data as in any of the preceding embodiments.
According to still further embodiments of the present disclosure, there is provided a non-transitory computer-readable storage medium having stored thereon a computer program, wherein the program, when executed by a processor, implements the long-term determination method of traffic data of any of the foregoing embodiments.
The method comprises the steps of dividing each first time period corresponding to the service data to be determined into a plurality of second time periods, wherein the first time periods are divided according to time units with fine granularity, and the second time periods are divided according to time units with coarse granularity. And then, for each second time period, determining the service data statistic value in the second time period according to the service data statistic value in each second time period in a plurality of second time periods before the second time period, and then for each first time period, determining the service data value in the first time period according to the service data statistic value in the second time period corresponding to the first time period, thereby obtaining the value of the service data to be determined in each first time period. As the trend and the characteristic presented in the coarse-grained time period are different from those presented in the fine-grained time period aiming at long-term service data, the trend and the characteristic are determined in different modes instead of one mode, and the accuracy of determining the service data is improved.
Other features of the present disclosure and advantages thereof will become apparent from the following detailed description of exemplary embodiments thereof, which proceeds with reference to the accompanying drawings.
Drawings
In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present disclosure, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 shows a flow diagram of a method of long-term determination of traffic data according to some embodiments of the present disclosure.
Fig. 2 shows a flow diagram of a method for long-term determination of traffic data according to further embodiments of the present disclosure.
Fig. 3 shows a schematic structural diagram of a long-term traffic data determination apparatus according to some embodiments of the present disclosure.
Fig. 4 shows a schematic structural diagram of a long-term traffic data determination apparatus according to another embodiment of the present disclosure.
Fig. 5 is a schematic structural diagram of a device for long-term determination of traffic data according to further embodiments of the present disclosure.
Detailed Description
The technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the drawings in the embodiments of the present disclosure, and it is obvious that the described embodiments are only a part of the embodiments of the present disclosure, and not all of the embodiments. The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the disclosure, its application, or uses. All other embodiments, which can be derived by a person skilled in the art from the embodiments disclosed herein without making any creative effort, shall fall within the protection scope of the present disclosure.
The present disclosure proposes a method for determining service data for a long time, which is described below with reference to fig. 1.
Fig. 1 is a flowchart of some embodiments of a long-term determination method of service data according to the present disclosure. As shown in fig. 1, the method of this embodiment includes: steps S102 to S106.
In step S102, each first time period corresponding to the service data to be determined is divided into one or more groups, and each group corresponds to a second time period.
The first time period and the second time period are divided according to time units with different granularities respectively, the first time period corresponds to a time unit with a fine granularity, and the second time period corresponds to a time unit with a coarse granularity. For example, where the first time period is an hour, the second time period may be a day; where the first time period is a day, the second time period may be a week or a month, and so on. The following examples take the first time period as a day and the second time period as a week. Each first time period corresponding to the service data to be determined is, for example, each of the days 1 to 91 in the future. According to the international time division mode, the conversion method of the granularity corresponding to the first time period and the granularity corresponding to the second time period divides each first time period into each second time period, for example, the first time period is divided into the first week to 13 weeks in the future from the 1 st to 91 th days, and if the first day in the future is not Monday and is not divided into the first week in the future, the first time period can be divided into the current week.
In step S104, for each second time period, the service data statistics value in each second time period in a plurality of second time periods before the second time period is determined according to the service data statistics value in the second time period.
The statistics of the service data in the second time period is, for example, an average, a sum, or a sum of squares of the service data in each of the first time periods in the second time period, and is not limited to the illustrated example. For example, for each week to be determined, the business data statistics value in the week is determined according to the business data statistics value in each week in a plurality of weeks before the week. In some embodiments, a plurality of second time periods within a preset time window length range before the second time period and the service data statistics value in each second time period are determined; and inputting the service data statistic value in each second time period into a pre-trained coarse-grained determination model to obtain the output service data statistic value in the second time period. The preset time window length may be trained. For example, the traffic data statistics for 24 consecutive second time periods prior to each second time period may be input into a pre-trained coarse-grained determination model.
In some embodiments, the coarse granularity determination model is used to represent the relationship between the traffic data statistics in one second time period and the traffic data statistics in each of a plurality of second time periods before the second time period. The coarse-grained determination model comprises the following steps: ridge regression models, timing models, or random forest models, are not limited to the examples given. For example, in the case where the coarse-grained determination model is a ridge regression model, the following formula may be employed to determine the traffic data statistics over the second time period.
In formula (1), y represents a service data statistic, lag, to be determined in the second time periodkRepresents the traffic data statistics, w, in the kth second time period before the second time period to be determinedkRepresenting parameters of the coarse-grained determination model, k being a positive integer, N representing a number of second time periods preceding the second time period to be determined. w is akIs obtained by pre-training. For example, the business data statistics for each week in the future 1 to 13 weeks can be predicted using the formula, for example, for the first week, the business data statistics for the previous 24 weeks before the first week can be used to input the coarse-grained determination model, for the second week, the business data statistics for the first week can be used, and for the previous 23 weeks before the first week, the coarse-grained determination model, and so on.
In step S106, for each first time period, the service data value of the first time period is determined according to the service data statistic value in the second time period corresponding to the first time period, so as to obtain the value of the service data to be determined of each first time period.
For example, for each day to be determined, the service data value of the day is determined according to the service data statistic value in the week corresponding to the day. In some embodiments, according to the time sequence of the first time period in the corresponding second time period, determining a pre-trained fine-grained determination model corresponding to the first time period; and inputting the business data statistic value in the second time period corresponding to the first time period into the fine-grained determination model corresponding to the first time period to obtain the output business data value of the first time period.
The parameters in the fine-grained determination models corresponding to the first time periods in different time sequences can be different, and the fine-grained determination models are used for representing the relationship between the business data values in the first time periods and the business data statistical values in the corresponding second time periods. For example, the parameters in the fine-grained models corresponding to monday through sunday may be different for each day. For example, the relationship between the service data value of the first time period and the service data statistic value in the second time period corresponding to the first time period may be represented by the following formula.
pi＝fi(P) (2)
In the formula (2), piThe method comprises the steps of representing a service data value of a first time period to be determined, i represents a time sequence of the first time period in a corresponding second time period, and P represents a service data statistic value in the second time period corresponding to the first time period. For example, p1Service data value, p, which may represent Monday6Service data value for Saturday according to different f1(P) and f6(P) is calculated. The fine-grained determination model is, for example, a random forest model or a ridge regression model, and is not limited to the examples given above, and the fine-grained determination model may be trained in advance according to the rules of actual data to determine the form of the fine-grained determination model.
In some cases, the earliest first time period corresponding to the service data to be determined does not belong to the first time period of the time sequence in the corresponding second time period, for example, the earliest first time period is wednesday, which is not the first day of the week, in this case, the service data values of the first time periods before the first time period in the same second time period are determined, which are not to be determined, and the service data statistics values in the second time period may be adjusted according to the determined service data values.
In some embodiments, when a first time period with the earliest time corresponding to service data to be determined does not belong to a first time period with a first time sequence in a corresponding second time period, the service data statistics value in the second time period corresponding to the first time period is adjusted according to the service data value of each first time period before the first time period belonging to the same second time period and the service data statistics value in the second time period corresponding to the first time period; and determining the business data value of the first time period according to the adjusted business data statistic value in the second time period corresponding to the first time period.
Further, the service data statistic value in a second time period corresponding to the first time period is adjusted according to the following formula:
in formula (3), P represents a statistical value of the service data in a second time period corresponding to the first time period, and PmThe method includes the steps that a service data value of an M-th first time period before a first time period belonging to the same second time period is represented, N represents the number of the first time periods contained in one second time period, M represents the number of the first time periods before the first time period belonging to the same second time period, and M is a positive integer. For example, if the first time period is Wednesday, then N is 7, M is 2, and p is known to be1，p2And calculating P' by using P obtained by the coarse granularity determination model.
And after the business data statistic value in the second time period corresponding to the first time period is adjusted, the adjusted business data statistic value in the second time period corresponding to the first time period is input into the fine-grained determination model corresponding to the first time period, and the output business data value of the first time period is obtained.
In the above embodiment, each first time period corresponding to the service data to be determined is divided into a plurality of second time periods, the first time periods are divided according to the time unit with the fine granularity, and the second time periods are divided according to the time unit with the coarse granularity. And then, for each second time period, determining the service data statistic value in the second time period according to the service data statistic value in each second time period in a plurality of second time periods before the second time period, and then for each first time period, determining the service data value in the first time period according to the service data statistic value in the second time period corresponding to the first time period, thereby obtaining the value of the service data to be determined in each first time period. As the trend and the characteristic presented in the coarse-grained time period are different from those presented in the fine-grained time period aiming at long-term service data, the trend and the characteristic are determined in different modes instead of one mode, and the accuracy of determining the service data is improved. The method comprises the steps that business data to be determined for a long time are determined on a coarse-grained time period, time intervals of the data are lengthened, long-term data features are easier to learn and more accurate to learn, a fine-grained model is further adopted for accurate determination, in a coarse-grained second time period, the business data show more time series characteristics, and in each fine-grained first time period in each second time period, the business data show more periodic characteristics. Therefore, according to the method of the above embodiment, the determination of the long-term service data is determined more accurately.
The inventor also verifies the accuracy of the scheme according to experiments. The method of the scheme is used for predicting the service data from 6 days 1 and 6 days 2014 to 29 days 6 and 6 months 2019, and the RMSE (root mean square error) of the prediction result and the actual data is much smaller than that of the prior art which only uses a time sequence model, a random forest model, a ridge regression model and the like.
In some embodiments, the traffic data includes: the price of the product, the sales volume of the product, the visit volume of the website, or the traffic volume of the website may be other data, and is not limited to the examples given. In the case that the service data is the price of the commodity, the sales volume of the commodity, or the access volume of the commodity, the commodity may be recommended to the user according to the value of the service data to be determined for each first time period. For example, the time at which the price is the lowest is determined from the price of the commodity determined for a long period of time, and the commodity is recommended to the relevant user at that time, or the commodity may be recommended to the relevant user at that time when the sales volume of the commodity is low or the access volume of the commodity is low.
And under the condition that the service data is the access volume of the website or the flow of the website, adjusting the resource deployment corresponding to the website according to the value of the service data to be determined in each first time period. For example, according to the determined access amount of the website or the determined flow of the website corresponding to each first time period, the time exceeding the access amount threshold or the time exceeding the flow threshold is determined, and the reminding information is sent. For another example, the determined website access amount or website traffic corresponding to each first time period is compared with a preset access amount range and a preset traffic amount range, and different access amount ranges and traffic amount ranges correspond to different deployment resource amounts, so that the deployment resource amounts of different first time periods are determined, and dynamic adjustment of resources is realized. The resource is, for example, one or more of a server, a bandwidth, a CPU, and a storage space.
Further embodiments of the disclosed method for long-term determination of traffic data are described below in conjunction with fig. 2.
Fig. 2 is a flowchart of another embodiment of a method for long-term determination of traffic data according to the present disclosure. As shown in fig. 2, the method of this embodiment includes: steps S202 to S206.
In step S202, historical service data values in each first time period and service data statistics values in each second time period are obtained.
The data can be acquired in a web crawler mode, and abnormal value elimination and missing value filling can be performed on the acquired data.
In some embodiments, for each traffic data value of the first time period, determining whether the traffic data value of the first time period is abnormal according to a mean value and a variance, deleting the traffic data value of the first time period in case that the traffic data value of the first time period is abnormal, and taking the traffic data value of the first time period which is previous to the first time period as the traffic data value of the first time period.
In step S204, the coarse-grained determination model is trained according to the historical traffic data statistical values in each second time period.
The coarse-grained determination model is, for example, a model corresponding to formula (1). The statistical values of the business data in each historical second time period can be input into the coarse granularity determination model to obtain an output result, loss is calculated according to the output result and the true value, and parameters of the model are adjusted according to the loss until the loss value is minimum. For example, the loss function can be expressed by the following formula.
In formula (4), y represents the service data statistic, lag, to be determined in the second time periodkRepresents the traffic data statistics, w, in the kth second time period before the second time period to be determinedkAnd parameters representing a coarse-grained determination model, k is a positive integer, and C is a regularization factor.
In step S206, a fine-grained determination model is trained according to historical traffic data values of each first time period and traffic data statistical values of each second time period.
For example, historical service data values of each first time period and service data statistics values of each second time period are input into a fine-grained determination model to obtain an output result, loss is calculated according to the output result and a true value, and parameters of the model are adjusted according to the loss until the loss value is minimum.
Step S204 and step S206 may be executed in parallel without being sequentially executed.
Some application examples of the present disclosure may include the following five stages. The first, business data extraction stage, extracting business data, preprocessing historical business data, eliminating abnormal values, filling missing values, and calculating the historical business data statistics value of each week. Second, a coarse-grained determination model, taking the determination of business data for the next 13 weeks (91 days) as an example, builds a machine learning model based on historical data. And thirdly, learning the relationship between the weekly business data statistic and the daily business data value through historical data to establish a fine-grained determination model. And fourthly, in the prediction stage, firstly, the business data statistic value of each week of 13 weeks in the future is predicted according to the coarse-grained determination model, and then the business data value of each day is obtained by inputting the business data statistic value of each week into the fine-grained determination model. And fifthly, in the business data value fine adjustment stage, if the predicted first week is not a complete week, fine adjustment can be carried out on the business data statistic value of the week by using the existing business data value.
The present disclosure also provides a device for determining long-term service data, which is described below with reference to fig. 3.
Fig. 3 is a block diagram of some embodiments of the disclosed device for long-term determination of traffic data. As shown in fig. 3, the apparatus 30 of this embodiment includes: a time division module 310, a coarse granularity determination module 320, and a fine granularity determination module 330.
The time dividing module 310 is configured to divide each first time period corresponding to the service data to be determined into one or more groups, where each group corresponds to a second time period; the first time period and the second time period are divided according to time units with different granularities.
The coarse granularity determining module 320 is configured to, for each second time period, determine a service data statistic value in each second time period in a plurality of second time periods before the second time period according to the service data statistic value in the second time period.
In some embodiments, the coarse granularity determining module 320 is configured to determine a plurality of second time periods within a preset time window length range before the second time period, and a traffic data statistic in each second time period; inputting the service data statistic value in each second time period into a pre-trained coarse-grained determination model to obtain an output service data statistic value in the second time period; wherein, the coarse-grained determination model comprises: a ridge regression model, a timing model, or a random forest model.
The fine-grained determination module 330 is configured to, for each first time period, determine a service data value of the first time period according to the service data statistic value in the second time period corresponding to the first time period, so as to obtain a value of service data to be determined of each first time period.
In some embodiments, the fine-grained determination module 330 is configured to determine a pre-trained fine-grained determination model corresponding to the first time period according to a time sequence of the first time period in a corresponding second time period; inputting the business data statistic value in a second time period corresponding to the first time period into a fine-grained determination model corresponding to the first time period to obtain an output business data value of the first time period; the fine-grained determination models corresponding to the first time periods in different time sequences have different parameters, and the fine-grained determination models are used for expressing the relation between the business data values of the first time periods and the business data statistical values in the corresponding second time periods.
In some embodiments, the fine-grained determination module 330 is configured to adjust the traffic data statistics in the second time period corresponding to the first time period according to the traffic data values in the first time periods before the first time period belonging to the same second time period and the traffic data statistics in the second time period corresponding to the first time period; wherein the traffic data value of each first time period before the first time period is determined; and determining the business data value of the first time period according to the adjusted business data statistic value in the second time period corresponding to the first time period.
In some embodiments, the service data statistics in the second time period corresponding to the first time period are adjusted according to the following formula:
wherein, P represents the business data statistic value in the second time period corresponding to the first time period, and PmThe value of the traffic data of the mth first time period before the first time period belonging to the same second time period is represented, N represents the number of the first time periods included in one second time period, and M represents the number of the first time periods before the first time period belonging to the same second time period.
In some embodiments, the fine-grained determination module 330 is configured to determine a pre-trained fine-grained determination model corresponding to the first time period according to a time sequence of the first time period in a corresponding second time period; inputting the adjusted business data statistic value in a second time period corresponding to the first time period into a fine-grained determination model corresponding to the first time period to obtain an output business data value of the first time period; the fine-grained determination models corresponding to the first time periods in different time sequences have different parameters, and the fine-grained determination models are used for expressing the relation between the business data values of the first time periods and the business data statistical values in the corresponding second time periods.
In some embodiments, the apparatus 30 further comprises: the training module 340 is configured to obtain historical service data values in each first time period and service data statistics values in each second time period; training the coarse-grained determination model according to the statistical value of the business data in each historical second time period; the coarse granularity determination model is used for representing the relation between the business data statistic value in one second time period and the business data statistic value in each second time period in a plurality of second time periods before the second time period; and training a fine-grained determination model according to the historical business data values of the first time periods and the historical business data statistical values of the second time periods, wherein the fine-grained determination model is used for expressing the relation between the business data values of the first time periods and the corresponding business data statistical values in the second time periods.
In some embodiments, the apparatus 30 further comprises: a data processing module 350, configured to determine an average value and a variance of the service data values in each historical first time period; determining whether the service data value of each first time period is abnormal or not according to the average value and the variance aiming at the service data value of each first time period; and under the condition that the service data value of the first time period is abnormal, deleting the service data value of the first time period, and taking the service data value of the first time period before the first time period as the service data value of the first time period.
In some embodiments, the traffic data includes: price of the commodity, sales volume of the commodity, visit volume of the website, or traffic volume of the website. In some embodiments, the apparatus 30 further comprises: the recommending module 360 is configured to recommend a commodity to the user according to the value of the to-be-determined service data in each first time period when the service data is the price of the commodity, the sales volume of the commodity, or the access volume of the commodity; or, the resource scheduling module 370 is configured to, when the service data is the access volume of the website or the traffic volume of the website, adjust resource deployment corresponding to the website according to the value of the service data to be determined in each first time period.
The long-term determination device of the business data in the embodiments of the present disclosure may be implemented by various computing apparatuses or computer systems, which are described below with reference to fig. 4 and 5.
Fig. 4 is a block diagram of some embodiments of the disclosed device for long-term determination of traffic data. As shown in fig. 4, the apparatus 40 of this embodiment includes: a memory 410 and a processor 420 coupled to the memory 410, the processor 420 configured to perform a method of long-term determination of traffic data in any of the embodiments of the present disclosure based on instructions stored in the memory 410.
Fig. 5 is a block diagram of another embodiment of the disclosed long-term service data determination apparatus. As shown in fig. 5, the apparatus 50 of this embodiment includes: memory 510 and processor 520 are similar to memory 410 and processor 420, respectively. An input output interface 530, a network interface 540, a storage interface 550, and the like may also be included. These interfaces 530, 540, 550 and the connections between the memory 510 and the processor 520 may be, for example, via a bus 560. The input/output interface 530 provides a connection interface for input/output devices such as a display, a mouse, a keyboard, and a touch screen. The network interface 540 provides a connection interface for various networking devices, such as a database server or a cloud storage server. The storage interface 550 provides a connection interface for external storage devices such as an SD card and a usb disk.
As will be appreciated by one skilled in the art, embodiments of the present disclosure may be provided as a method, system, or computer program product. Accordingly, the present disclosure may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present disclosure may take the form of a computer program product embodied on one or more computer-usable non-transitory storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present disclosure is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the disclosure. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above description is only exemplary of the present disclosure and is not intended to limit the present disclosure, so that any modification, equivalent replacement, or improvement made within the spirit and principle of the present disclosure should be included in the scope of the present disclosure.
Claims (13)
1. A long-term determination method of service data comprises the following steps:
dividing each first time period corresponding to the service data to be determined into one or more groups, wherein each group corresponds to a second time period; the first time period and the second time period are divided according to time units with different granularities respectively;
for each second time period, determining the service data statistic value in each second time period in a plurality of second time periods before the second time period according to the service data statistic value in the second time period;
and for each first time period, determining the service data value of the first time period according to the service data statistic value in the second time period corresponding to the first time period, thereby obtaining the value of the service data to be determined of each first time period.
2. The long-term determination method of traffic data according to claim 1,
for each second time period, determining the service data statistic value in each second time period in a plurality of second time periods before the second time period according to the service data statistic value in each second time period comprises:
determining a plurality of second time periods within a preset time window length range before the second time period and a service data statistic value in each second time period;
inputting the service data statistic value in each second time period into a pre-trained coarse-grained determination model to obtain an output service data statistic value in the second time period;
wherein the coarse-grained determination model comprises: a ridge regression model, a timing model, or a random forest model.
3. The long-term determination method of traffic data according to claim 1,
for each first time period, determining the service data value of the first time period according to the service data statistic value in the second time period corresponding to the first time period includes:
according to the time sequence of the first time period in the corresponding second time period, determining a pre-trained fine-grained determination model corresponding to the first time period;
inputting the business data statistic value in a second time period corresponding to the first time period into a fine-grained determination model corresponding to the first time period to obtain an output business data value of the first time period;
the fine-grained determination models corresponding to the first time periods in different time sequences have different parameters, and the fine-grained determination models are used for expressing the relation between the business data values of the first time periods and the business data statistical values in the corresponding second time periods.
4. The long-term determination method of traffic data according to claim 1,
for each first time period, determining the service data value of the first time period according to the service data statistic value in the second time period corresponding to the first time period includes:
adjusting the service data statistics value in the second time period corresponding to the first time period according to the service data value in each first time period before the first time period belonging to the same second time period and the service data statistics value in the second time period corresponding to the first time period; wherein the traffic data value of each first time period before the first time period is determined;
and determining the business data value of the first time period according to the adjusted business data statistic value in the second time period corresponding to the first time period.
5. The long-term determination method of traffic data according to claim 4,
and adjusting the service data statistic value in a second time period corresponding to the first time period according to the following formula:
wherein, P represents the business data statistic value in the second time period corresponding to the first time period, and PmThe value of the traffic data of the mth first time period before the first time period belonging to the same second time period is represented, N represents the number of the first time periods included in one second time period, and M represents the number of the first time periods before the first time period belonging to the same second time period.
6. The long-term determination method of traffic data according to claim 4,
the determining the service data value of the first time period according to the adjusted service data statistic value in the second time period corresponding to the first time period includes:
according to the time sequence of the first time period in the corresponding second time period, determining a pre-trained fine-grained determination model corresponding to the first time period;
inputting the adjusted business data statistic value in a second time period corresponding to the first time period into a fine-grained determination model corresponding to the first time period to obtain an output business data value of the first time period;
the fine-grained determination models corresponding to the first time periods in different time sequences have different parameters, and the fine-grained determination models are used for expressing the relation between the business data values of the first time periods and the business data statistical values in the corresponding second time periods.
7. The long-term determination method of traffic data according to claim 1, further comprising:
obtaining historical business data values of all first time periods and business data statistics values of all second time periods;
training the coarse-grained determination model according to the statistical value of the business data in each historical second time period; the coarse granularity determination model is used for representing the relation between the business data statistic value in one second time period and the business data statistic value in each second time period in a plurality of second time periods before the second time period;
and training a fine-grained determination model according to the historical business data values of the first time periods and the historical business data statistical values of the second time periods, wherein the fine-grained determination model is used for expressing the relation between the business data values of the first time periods and the corresponding business data statistical values in the second time periods.
8. The long-term determination method of traffic data according to claim 7, further comprising:
determining the average value and the variance of the business data values of each historical first time period;
determining whether the service data value of each first time period is abnormal or not according to the average value and the variance aiming at the service data value of each first time period;
and under the condition that the service data value of the first time period is abnormal, deleting the service data value of the first time period, and taking the service data value of the first time period before the first time period as the service data value of the first time period.
9. The long-term determination method of traffic data according to any one of claims 1-8,
the service data comprises: price of the commodity, sales volume of the commodity, visit volume of the website, or traffic volume of the website.
10. The long-term determination method of traffic data according to claim 9, further comprising:
recommending commodities to the user according to the value of the to-be-determined business data of each first time period under the condition that the business data are the price of the commodities, the sales volume of the commodities or the access volume of the commodities;
or, when the service data is the access volume of the website or the flow of the website, adjusting the resource deployment corresponding to the website according to the value of the service data to be determined in each first time period.
11. An apparatus for long-term determination of traffic data, comprising:
the time division module is used for dividing each first time period corresponding to the service data to be determined into one or more groups, and each group corresponds to a second time period; the first time period and the second time period are divided according to time units with different granularities respectively;
the coarse granularity determining module is used for determining a service data statistic value in each second time period in a plurality of second time periods before the second time period according to the service data statistic value in each second time period;
and the fine granularity determining module is used for determining the business data value of each first time period according to the business data statistic value in the second time period corresponding to the first time period so as to obtain the value of the business data to be determined of each first time period.
12. An apparatus for long-term determination of traffic data, comprising:
a processor; and
a memory coupled to the processor for storing instructions that, when executed by the processor, cause the processor to perform a method of long-term determination of traffic data according to any of claims 1-10.
13. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the program when executed by a processor implements the steps of the method of any one of claims 1-10.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010091464.3A CN111339156B (en) | 2020-02-07 | 2020-02-07 | Method, apparatus and computer readable storage medium for long-term determination of business data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010091464.3A CN111339156B (en) | 2020-02-07 | 2020-02-07 | Method, apparatus and computer readable storage medium for long-term determination of business data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111339156A true CN111339156A (en) | 2020-06-26 |
| CN111339156B CN111339156B (en) | 2023-09-26 |
Family
ID=71183580
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010091464.3A Active CN111339156B (en) | 2020-02-07 | 2020-02-07 | Method, apparatus and computer readable storage medium for long-term determination of business data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111339156B (en) |
Citations (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2488079A1 (en) * | 2003-11-20 | 2005-05-20 | New England 800 Company D/B/A Taction | System and method for event-based forecasting |
| US20110251878A1 (en) * | 2010-04-13 | 2011-10-13 | Yahoo! Inc. | System for processing large amounts of data |
| US20140081652A1 (en) * | 2012-09-14 | 2014-03-20 | Risk Management Solutions Llc | Automated Healthcare Risk Management System Utilizing Real-time Predictive Models, Risk Adjusted Provider Cost Index, Edit Analytics, Strategy Management, Managed Learning Environment, Contact Management, Forensic GUI, Case Management And Reporting System For Preventing And Detecting Healthcare Fraud, Abuse, Waste And Errors |
| CN104252647A (en) * | 2014-06-10 | 2014-12-31 | 南瑞（武汉）电气设备与工程能效测评中心 | Inverse distance weighting interpolation method-based electrical load prediction method |
| CN104424598A (en) * | 2013-09-06 | 2015-03-18 | 株式会社日立制作所 | Cash demand quantity predicating device and method |
| CN105096067A (en) * | 2014-04-29 | 2015-11-25 | 阿里巴巴集团控股有限公司 | Logistics network early warning information generation method and server |
| CN105392154A (en) * | 2014-09-05 | 2016-03-09 | 中兴通讯股份有限公司 | Resource occupation prediction method and system |
| WO2016095708A1 (en) * | 2014-12-16 | 2016-06-23 | 高德软件有限公司 | Traffic flow prediction method, and prediction model generation method and device |
| CN105844340A (en) * | 2015-02-03 | 2016-08-10 | 华为技术有限公司 | Data processing method and apparatus |
| CN107329887A (en) * | 2016-04-28 | 2017-11-07 | 腾讯科技（深圳）有限公司 | A kind of data processing method and device based on commending system |
| CN107958307A (en) * | 2017-11-28 | 2018-04-24 | 珠海格力电器股份有限公司 | Electricity charge prediction method and device |
| CN108491960A (en) * | 2018-02-11 | 2018-09-04 | 广东电网有限责任公司佛山供电局 | A kind of power distribution network power failure optimal case appraisal procedure |
| CN108880931A (en) * | 2018-05-29 | 2018-11-23 | 北京百度网讯科技有限公司 | Method and apparatus for output information |
| CN109035761A (en) * | 2018-06-25 | 2018-12-18 | 复旦大学 | Travel time estimation method based on back-up surveillance study |
| CN109509017A (en) * | 2018-09-27 | 2019-03-22 | 中国平安人寿保险股份有限公司 | User's retention ratio prediction technique and device based on big data analysis |
| CN109726872A (en) * | 2018-12-29 | 2019-05-07 | 华润电力技术研究院有限公司 | A kind of energy consumption prediction technique, device, storage medium and electronic equipment |
| CN109993645A (en) * | 2017-12-29 | 2019-07-09 | 北京京东尚科信息技术有限公司 | Financial forecast method, system, computer system and computer readable storage medium |
| CN110009384A (en) * | 2019-01-07 | 2019-07-12 | 阿里巴巴集团控股有限公司 | Predict the method and device of operational indicator |
| CN110275814A (en) * | 2019-06-28 | 2019-09-24 | 深圳前海微众银行股份有限公司 | A monitoring method and device for a business system |
| CN110321240A (en) * | 2019-06-28 | 2019-10-11 | 阿里巴巴集团控股有限公司 | A kind of business impact assessment method and apparatus based on time series forecasting |
| CN110400005A (en) * | 2019-06-28 | 2019-11-01 | 阿里巴巴集团控股有限公司 | Time Series Forecasting Methods, device and the equipment of operational indicator |
-
2020
- 2020-02-07 CN CN202010091464.3A patent/CN111339156B/en active Active
Patent Citations (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2488079A1 (en) * | 2003-11-20 | 2005-05-20 | New England 800 Company D/B/A Taction | System and method for event-based forecasting |
| US20110251878A1 (en) * | 2010-04-13 | 2011-10-13 | Yahoo! Inc. | System for processing large amounts of data |
| US20140081652A1 (en) * | 2012-09-14 | 2014-03-20 | Risk Management Solutions Llc | Automated Healthcare Risk Management System Utilizing Real-time Predictive Models, Risk Adjusted Provider Cost Index, Edit Analytics, Strategy Management, Managed Learning Environment, Contact Management, Forensic GUI, Case Management And Reporting System For Preventing And Detecting Healthcare Fraud, Abuse, Waste And Errors |
| CN104424598A (en) * | 2013-09-06 | 2015-03-18 | 株式会社日立制作所 | Cash demand quantity predicating device and method |
| CN105096067A (en) * | 2014-04-29 | 2015-11-25 | 阿里巴巴集团控股有限公司 | Logistics network early warning information generation method and server |
| CN104252647A (en) * | 2014-06-10 | 2014-12-31 | 南瑞（武汉）电气设备与工程能效测评中心 | Inverse distance weighting interpolation method-based electrical load prediction method |
| CN105392154A (en) * | 2014-09-05 | 2016-03-09 | 中兴通讯股份有限公司 | Resource occupation prediction method and system |
| WO2016095708A1 (en) * | 2014-12-16 | 2016-06-23 | 高德软件有限公司 | Traffic flow prediction method, and prediction model generation method and device |
| CN105844340A (en) * | 2015-02-03 | 2016-08-10 | 华为技术有限公司 | Data processing method and apparatus |
| CN107329887A (en) * | 2016-04-28 | 2017-11-07 | 腾讯科技（深圳）有限公司 | A kind of data processing method and device based on commending system |
| CN107958307A (en) * | 2017-11-28 | 2018-04-24 | 珠海格力电器股份有限公司 | Electricity charge prediction method and device |
| CN109993645A (en) * | 2017-12-29 | 2019-07-09 | 北京京东尚科信息技术有限公司 | Financial forecast method, system, computer system and computer readable storage medium |
| CN108491960A (en) * | 2018-02-11 | 2018-09-04 | 广东电网有限责任公司佛山供电局 | A kind of power distribution network power failure optimal case appraisal procedure |
| CN108880931A (en) * | 2018-05-29 | 2018-11-23 | 北京百度网讯科技有限公司 | Method and apparatus for output information |
| CN109035761A (en) * | 2018-06-25 | 2018-12-18 | 复旦大学 | Travel time estimation method based on back-up surveillance study |
| CN109509017A (en) * | 2018-09-27 | 2019-03-22 | 中国平安人寿保险股份有限公司 | User's retention ratio prediction technique and device based on big data analysis |
| CN109726872A (en) * | 2018-12-29 | 2019-05-07 | 华润电力技术研究院有限公司 | A kind of energy consumption prediction technique, device, storage medium and electronic equipment |
| CN110009384A (en) * | 2019-01-07 | 2019-07-12 | 阿里巴巴集团控股有限公司 | Predict the method and device of operational indicator |
| CN110275814A (en) * | 2019-06-28 | 2019-09-24 | 深圳前海微众银行股份有限公司 | A monitoring method and device for a business system |
| CN110321240A (en) * | 2019-06-28 | 2019-10-11 | 阿里巴巴集团控股有限公司 | A kind of business impact assessment method and apparatus based on time series forecasting |
| CN110400005A (en) * | 2019-06-28 | 2019-11-01 | 阿里巴巴集团控股有限公司 | Time Series Forecasting Methods, device and the equipment of operational indicator |
Non-Patent Citations (2)
| Title |
|---|
| FANGCHAO LIU;JIANXIN LI;JICHUN CHEN;DONGSHENG HU;YING LI;JIANFENG HUANG;XIAOQING LIU;XUELI YANG;JIE CAO;CHONG SHEN;LING YU;ZHENDON: "Predicting lifetime risk for developing atherosclerotic cardiovascular disease in Chinese population:the China-PAR project", SCIENCE BULLETIN, no. 12 * |
| 张晚笛;陈峰;王子甲;汪波;王挺;: "基于多时间粒度的地铁出行规律相似性度量", 铁道学报, no. 04 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111339156B (en) | 2023-09-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109495318B (en) | Mobile communication network flow prediction method, device and readable storage medium | |
| US11250449B1 (en) | Methods for self-adaptive time series forecasting, and related systems and apparatus | |
| CN110443657B (en) | Client flow data processing method and device, electronic equipment and readable medium | |
| CN110019401B (en) | Method, device, equipment and storage medium for predicting part quantity | |
| US9336493B2 (en) | Systems and methods for clustering time series data based on forecast distributions | |
| KR20210017342A (en) | Time series prediction method and apparatus based on past prediction data | |
| CN109961165B (en) | Piece quantity forecasting method, device, equipment and storage medium thereof | |
| JP6547070B2 (en) | Method, device and computer storage medium for push information coarse selection sorting | |
| CN111340244B (en) | Prediction method, training method, apparatus, server and medium | |
| US10324705B2 (en) | System and method for run-time update of predictive analytics system | |
| EP3961384B1 (en) | Automatic derivation of software engineering artifact attributes from product or service development concepts | |
| WO2022126961A1 (en) | Method for target object behavior prediction of data offset and related device thereof | |
| US20240311650A1 (en) | Technologies for using machine learning models to assess time series data | |
| US11651271B1 (en) | Artificial intelligence system incorporating automatic model updates based on change point detection using likelihood ratios | |
| CN107357764B (en) | Data analysis method, electronic device, and computer storage medium | |
| CN112990583A (en) | Method and equipment for determining mold entering characteristics of data prediction model | |
| CN113159934A (en) | Method and system for predicting passenger flow of network, electronic equipment and storage medium | |
| CN108416619B (en) | Consumption interval time prediction method and device and readable storage medium | |
| US20240202579A1 (en) | Time series prediction execution based on deviation risk evaluation | |
| CN107644042B (en) | Software program click rate pre-estimation sorting method and server | |
| CN115204278A (en) | Abnormal sample detection method and device, electronic device and storage medium | |
| CN111339156B (en) | Method, apparatus and computer readable storage medium for long-term determination of business data | |
| CN108804640B (en) | Data grouping method, device, storage medium and equipment based on maximized IV | |
| CN109145207B (en) | Information personalized recommendation method and device based on classification index prediction | |
| CN113537631B (en) | Forecasting methods, devices, electronic equipment and storage media for drug demand |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |