CN113469394A - Information processing apparatus, information processing method, and computer-readable storage medium - Google Patents

Information processing apparatus, information processing method, and computer-readable storage medium Download PDF

Info

Publication number
CN113469394A
CN113469394A CN202010237421.1A CN202010237421A CN113469394A CN 113469394 A CN113469394 A CN 113469394A CN 202010237421 A CN202010237421 A CN 202010237421A CN 113469394 A CN113469394 A CN 113469394A
Authority
CN
China
Prior art keywords
model
resource
prediction
unit period
next unit
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
Application number
CN202010237421.1A
Other languages
Chinese (zh)
Inventor
杨铭
夏迎炬
马军
刘汝杰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fujitsu Ltd
Original Assignee
Fujitsu Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Fujitsu Ltd filed Critical Fujitsu Ltd
Priority to CN202010237421.1A priority Critical patent/CN113469394A/en
Publication of CN113469394A publication Critical patent/CN113469394A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06315—Needs-based resource requirements planning or analysis

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Theoretical Computer Science (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Game Theory and Decision Science (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Marketing (AREA)
  • Development Economics (AREA)
  • Data Mining & Analysis (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Educational Administration (AREA)
  • Artificial Intelligence (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The application discloses an information processing apparatus, an information processing method, and a computer-readable storage medium. The information processing apparatus includes: a history data acquisition unit configured to acquire history data on the resource, the history data including at least a required amount of each part included in the resource in a plurality of unit periods included in a predetermined history period; a prediction result acquisition unit configured to: for each part contained in the resource, generating a first group of prediction results related to the demand of the part in the next unit time period through a first group of prediction models, and generating a second group of prediction results related to the demand of the part in the next unit time period through a second group of prediction models; and a fusion unit configured to obtain, for each part included in the resource, a final predicted required amount of the part in a next unit period by fusing the first group of the prediction results and the second group of the prediction results.

Description

Information processing apparatus, information processing method, and computer-readable storage medium
Technical Field
The present disclosure relates to the field of information processing, and in particular, to an information processing apparatus, an information processing method, and a computer-readable storage medium.
Background
The forecast relating to the demand of the resource enables, for example, the resource provider to know in advance the number of the respective parts of the parts included in the resource which are demanded in the future, thereby preparing the respective parts included in the resource.
Disclosure of Invention
The following presents a simplified summary of the disclosure in order to provide a basic understanding of some aspects of the disclosure. However, it should be understood that this summary is not an exhaustive overview of the disclosure. It is not intended to identify key or critical elements of the disclosure or to delineate the scope of the disclosure. Its sole purpose is to present some concepts of the disclosure in a simplified form as a prelude to the more detailed description that is presented later.
An object of the present disclosure is to provide an improved information processing apparatus, information processing method, and computer-readable storage medium.
According to an aspect of the present disclosure, there is provided an information processing apparatus including: a history data acquisition unit configured to acquire history data on a resource containing at least one portion, the history data including at least a required amount of each portion contained in the resource in a plurality of unit periods respectively included during a predetermined history time period; a prediction result acquisition unit configured to: generating, for each portion of the resource content, a first set of predictions relating to the demand of that portion for the next unit period after the predetermined historical period of time by means of a first set of prediction models; and for each portion contained by the resource, generating a second set of prediction results related to the demand of the portion in the next unit time period after the predetermined historical time period through a second set of prediction models; and a fusion unit configured to obtain, for each part included in the resource, a final predicted required amount of the part in a next unit period after the predetermined historical period by fusing the first and second sets of prediction results, wherein the first set of prediction models is suitable for short-term prediction and the second set of prediction models is suitable for long-term prediction.
According to another aspect of the present disclosure, there is provided an information processing method including: a history data acquisition step of acquiring history data on a resource containing at least one portion, the history data including at least a required amount of each portion contained in the resource in a plurality of unit periods respectively included in a predetermined history time period; a prediction result acquisition step configured to: generating, for each portion of the resource content, a first set of predictions relating to the demand of that portion for the next unit period after the predetermined historical period of time by means of a first set of prediction models; and for each portion contained by the resource, generating a second set of prediction results related to the demand of the portion in the next unit time period after the predetermined historical time period through a second set of prediction models; and a fusion step configured to obtain, for each part included in the resource, a final predicted required amount of the part in a next unit period after the predetermined historical period by fusing the first and second sets of prediction results, wherein the first set of prediction models is suitable for short-term prediction and the second set of prediction models is suitable for long-term prediction.
According to other aspects of the present disclosure, there are also provided computer program code and a computer program product for implementing the above-described method according to the present disclosure, and a computer readable storage medium having recorded thereon the computer program code for implementing the above-described method according to the present disclosure.
Additional aspects of the disclosed embodiments are set forth in the description section that follows, wherein the detailed description is presented to fully disclose the preferred embodiments of the disclosed embodiments without imposing limitations thereon.
Drawings
The disclosure may be better understood by reference to the following detailed description taken in conjunction with the accompanying drawings, in which like or similar reference numerals are used throughout the figures to designate like or similar components. The accompanying drawings, which are incorporated in and form a part of the specification, further illustrate preferred embodiments of the present disclosure and explain the principles and advantages of the present disclosure, are incorporated in and form a part of the specification. Wherein:
fig. 1 is a block diagram showing a functional configuration example of an information processing apparatus according to an embodiment of the present disclosure;
FIG. 2 illustrates an example of a first set of predictive models and a second set of predictive models;
FIG. 3 illustrates an example for processing historical data to obtain processed historical data for a model 2;
fig. 4 shows an example of the 2 nd predicted required amount in the next unit period of each part of the acquisition resource inclusion in the example case shown in fig. 2;
FIG. 5 illustrates an example of a training process of a regression model in the case where the unit period is a month;
FIG. 6 is a flow chart illustrating an example of a flow of an information processing method 600 according to an embodiment of the present disclosure; and
fig. 7 is a block diagram showing an example structure of a personal computer employable in the embodiments of the present disclosure.
Detailed Description
Exemplary embodiments of the present disclosure will be described hereinafter with reference to the accompanying drawings. In the interest of clarity and conciseness, not all features of an actual implementation are described in the specification. It will of course be appreciated that in the development of any such actual embodiment, numerous implementation-specific decisions must be made to achieve the developers' specific goals, such as compliance with system-related and business-related constraints, which will vary from one implementation to another. Moreover, it will be appreciated that such a development effort might be complex and time-consuming, but would nevertheless be a routine undertaking for those of ordinary skill in the art having the benefit of this disclosure.
Here, it should be further noted that, in order to avoid obscuring the present disclosure with unnecessary details, only the device structures and/or processing steps closely related to the scheme according to the present disclosure are shown in the drawings, and other details not so relevant to the present disclosure are omitted.
Embodiments according to the present disclosure are described in detail below with reference to the accompanying drawings.
First, a functional configuration example of an information processing apparatus according to an embodiment of the present disclosure will be described with reference to fig. 1. Fig. 1 is a block diagram showing a functional configuration example of an information processing apparatus according to an embodiment of the present disclosure. As shown in fig. 1, the information processing apparatus 100 according to the embodiment of the present disclosure may include a history data acquisition unit 102, a prediction result acquisition unit 104, and a fusion unit 106.
The history data acquisition unit 102 may be configured to acquire history data on a resource, wherein the resource contains at least one portion, and the history data includes at least a required amount of each portion contained by the resource respectively in a plurality of unit periods included in a predetermined history time period. For example, the unit period may be a season, a month, a week, or the like.
The prediction result acquisition unit 104 may be configured to: generating, for each portion of the resource inclusion, a first set of predictions relating to the demand of that portion for the next unit period after a predetermined historical period of time by means of a first set of prediction models; and for each portion contained by the resource, generating a second set of predictions relating to the demand for that portion for the next unit period after the predetermined historical period by a second set of prediction models. Wherein the first set of prediction models is adapted for short-term prediction and the second set of prediction models is adapted for long-term prediction. By way of illustration and not limitation, the short-term prediction accuracy of the first set of prediction models is higher than the second set of prediction models, and the long-term prediction accuracy of the second set of prediction models is higher than the first set of prediction models. As a further example, the interference rejection capabilities of the second set of predictive models are greater than the interference rejection capabilities of the first set of predictive models.
The fusion unit 106 may be configured to obtain, for each part included in the resource, a final predicted required amount of the part in a next unit period after the predetermined history period by fusing the first and second sets of the prediction results.
By way of illustration and not limitation, resources may include, but are not limited to: various services, for example, regular maintenance services for vehicles such as vehicles, ships, airplanes, etc.; and various articles, for example, components included in vehicles such as vehicles, ships, airplanes, and the like; components included in electronic devices such as smart phones, portable computers, desktop computers, and the like.
As an example, the first set of predictive models may include 1 st through kth models, where k is a positive integer greater than 1, and the second set of predictive models may include k +1 st through 2k-1 st models. For example, the 1 st model included in the first set of predictive models may be obtained by training using historical data, and the ith model included in the first set of predictive models may be obtained by training using processed historical data as follows: the processed historical data is obtained by summing or averaging, for each portion of the resource containing, the demand of that portion in the historical data over i adjacent unit periods, where i is 2. The second set of predictive models may include a k +1 th model through a 2k-1 th model each established based on historical data. The first set of predictors may include: 1 st predicted demand amount in the next unit period after a predetermined history period for each part contained in the resource obtained by the 1 st model; and an ith estimated demand amount for each part contained in the resource obtained by the ith model, a next unit period after the predetermined history period, and i-1 unit periods immediately before the next unit period. The second set of predictors may include: each part contained in the resource obtained by the (i + k-1) th model included in the second group of prediction models, the (i + k-1) th predicted demand amount in the next unit period after the predetermined history period, and the (i + k-1) th estimated demand amount in the i-1 th unit period immediately before the next unit period.
The fusion unit 106 may be further configured to: for each part contained in the resource, obtaining an ith predicted demand quantity of the part in the next unit period based on a ratio of an i + k-1 th predicted demand quantity of the part in the next unit period, obtained through an i + k-1 th model, to an i + k-1 th estimated demand quantity in i-1 unit periods immediately before the next unit period, and an ith estimated demand quantity of the part in the next unit period and i-1 unit periods immediately before the next unit period, obtained through the i < th > model; and for each part contained in the resource, obtaining the final predicted demand of the part based on the 1 st predicted demand to the 2k-1 st predicted demand.
As an example, the fusion unit 106 may obtain, for each part of the resource inclusion, a mean value or a median value of the 1 st to 2k-1 st predicted demand amounts as a final predicted demand amount of the part. In addition, those skilled in the art may obtain the final predicted required amount of each part in other manners based on the 1 st to 2k-1 th predicted required amounts of each part, for example, obtain an average value of values except for the maximum value and the minimum value among the 1 st to 2k-1 th predicted required amounts as the final predicted required amount, and so on, which will not be described herein again.
Fig. 2 shows an example of the first and second sets of prediction models in the case where k is 4. In fig. 2, for the history data, for example, each block represents a demand amount in one unit period. As shown in fig. 2, the first set of predictive models includes a1 st model, a2 nd model, a3 rd model, and a4 th model, and the second set of predictive models includes a5 th model, a6 th model, and a7 th model. As described above, the 1 st model may be obtained by training using historical data. For clarity of illustration, the first and second sets of predictive models are not shown in FIG. 2, but rather the processing performed by each model in each set is shown. The 2 nd model may be obtained by training with processed historical data as follows: the processed history data is obtained by summing or averaging, for each portion of the resource inclusion, the demand amounts of the portion in the history data in adjacent 2 unit periods. The 3 rd model may be obtained by training with processed historical data as follows: the processed history data is obtained by summing or averaging, for each portion of the resource inclusion, the demanded amounts of the portion in the history data in adjacent 3 unit periods. The 4 th model may be obtained by training with processed historical data as follows: the processed history data is obtained by summing or averaging, for each part of the resource inclusion, the demanded amounts of the part in the history data over 4 unit periods. The processed history data for the 2 nd model, the 3 rd model, and the 4 th model may be referred to as second processed history data, third processed history data, and fourth processed history data, respectively. In fig. 2, for example, for the second processed history data, each block may represent a result obtained by summing or averaging the demand amounts in adjacent 2 unit periods; for the third processed history data, each block may represent a result obtained by summing or averaging the demand amounts in adjacent 3 unit periods; for the fourth processed history data, each block may represent a result obtained by summing or averaging the demand amounts in adjacent 4 unit periods. Further, each of the 5 th, 6 th and 7 th models may be built based on historical data.
FIG. 3 shows an example for processing historical data to obtain processed historical data for model 2. As shown in fig. 3, the demand in adjacent 2 unit periods is summed or averaged to obtain the processed historical data for model 2.
It is noted that, although it is described above that the demanded quantities in the adjacent unit time periods are summed or averaged to obtain the processed historical data for the corresponding model, a person skilled in the art may process the demanded quantities in the adjacent unit time periods in other manners to obtain the corresponding processed historical data, which will not be described herein again.
Referring again to fig. 2, the first set of predictors may include: 1 st predicted demand a1 for the next unit period (i.e., m-th unit period) for each part contained in the resource obtained by the 1 st model; and 2 nd estimated demand B2 for each part contained in the resource obtained by the 2 nd model, in the next unit period and 1 unit period (i.e., m-1 th and m unit periods) immediately before the next unit period; a3 rd estimated demand B3 for each part contained in the resource obtained by the 3 rd model, in the next unit period and 2 unit periods (i.e., m-2, m-1, and m unit periods) immediately before the next unit period; and a4 th estimated demand B4 for each part contained in the resource obtained by the 4 th model, in the next unit period and 3 unit periods (i.e., m-3, m-2, m-1, and m unit periods) immediately before the next unit period. The second set of predictors may include: each part contained in the resource obtained by the 5 th model, the 5 th predicted demand a5 in the next unit period (i.e., the m-th unit period), and the 5 th estimated demand B5 in 1 unit period (i.e., the m-1 th unit period) immediately before the next unit period; each section contained in the resource obtained by the 6 th model, the 6 th predicted demand a6 in the next unit period (i.e., the m-th unit period), and the 6 th estimated demand B6 in 2 unit periods (i.e., the m-2 th and m-1 th unit periods) immediately before the next unit period; each part contained in the resource obtained by the 7 th model, the 7 th predicted demand a7 in the next unit period (i.e., the m-th unit period), and the 7 th estimated demand B7 in 3 unit periods (i.e., the m-3, m-2, and m-1 unit periods) immediately before the next unit period.
It is to be noted that, in fig. 2, for the prediction by the 1 st model, although arrow (r) is shown pointing from the m-1 st unit period to the m-th unit period, this does not mean that only the required amount in the m-1 st unit period is used as an input to the 1 st model to obtain the 1 st predicted required amount a 1. Actually, the unit period (i.e., the m-1 st unit period) corresponding to the start point of the arrow (r) represents the cutoff point of the history data used for obtaining the 1 st predicted required amount a 1. That is, those skilled in the art can select, as inputs to the 1 st model, the demand amounts in one or more unit time periods in the unit time period corresponding to the start point of the arrow (r) and the unit time periods before the start point of the arrow (r) (i.e., the 1 st to m-2 th unit time periods) according to actual needs to obtain the 1 st predicted demand amount a 1. The same applies to the 2 nd to 7 th models. For example, for the 2 nd model, one or more processed history data among the processed history data corresponding to the start point of the arrow (c) and the processed history data before the start point of the arrow (c) may be used as an input of the 2 nd model to obtain the 2 nd estimated demand B2. Further, for example, for the 5 th model, the demand amounts in one or more unit periods of the unit period (i.e., the m-2 th unit period) corresponding to the start point of the arrow (c) and the unit periods before the start point of the arrow (c) (i.e., the 1 st to m-2 th unit periods) may be used as inputs of the 5 th model to obtain the 5 th predicted demand amount a5 and the 5 th estimated demand amount B5. In addition, the unit periods corresponding to the start point of the arrow (c) and the required amounts in all the unit periods before the start point of the arrow (c) may also be used as inputs of the 5 th model to obtain the 5 th predicted required amount a5 and the 5 th estimated required amount B5.
Fig. 4 shows an example of the 2 nd predicted required amount in the next unit period of each part of the acquisition resource inclusion in the example case shown in fig. 2. As shown in fig. 4, the fusion unit 106 may obtain, for each section of the resource inclusion, a2 nd predicted demand amount a2 of the section in the next unit period based on a ratio a5/B5 of a5 th predicted demand amount a5 of the section in the next unit period (i.e., the m-th unit period) obtained by the 5 th model and a5 th estimated demand amount B5 of the section in 1 unit period immediately before the next unit period (i.e., the m-1 th unit period), and a2 nd estimated demand amount B2 of the section in the next unit period and 1 unit period immediately before the next unit period (i.e., the m-1 th and m-unit periods) obtained by the 2 nd model. For example, the fusion unit 106 may multiply, for each section that the resource contains, the ratio a5/B5 of the 5 th predicted demand amount a5 and the 5 th estimated demand amount B5 of the section by the 2 nd estimated demand amount B2 of the section, and take the obtained result as the 2 nd predicted demand amount a2 of the section. Further, the fusion unit 106 may obtain the 3 rd predicted required amount A3 and the 4 th predicted required amount a4 of each part included in the resource in a similar manner, and obtain the final predicted required amount of each part based on the 1 st predicted required amount a1 to the 7 th predicted required amount a7 of the part.
Note that, in practice, the true values (i.e., the actual demand amounts) corresponding to the 5 th estimated demand amount, the 6 th estimated demand amount, and the 7 th estimated demand amount are known, whereas the estimated demand amounts (i.e., the 5 th estimated demand amount B5, the 6 th estimated demand amount B6, and the 7 th estimated demand amount B7) obtained by the 5 th predictive model, the 6 th predictive model, and the 7 th predictive model are generally used in calculating the 2 nd predicted demand amount a2, the 3 rd predicted demand amount A3, and the 4 th predicted demand amount a4, respectively.
As an example, the second group of prediction models may further include a2 k-th model established based on the history data, and the second group of prediction results may further include a2 k-th predicted required amount in the next unit period for each part of the resource inclusion obtained by the 2 k-th model. In this case, the fusion unit 106 may obtain the final predicted required amount of each section based on the 1 st to 2k th predicted required amounts of the section.
Note that, although in the above-described example of the present embodiment, the first group of prediction models includes the 1 st model to the kth model, and the second group of prediction models includes the k +1 th model to the 2k-1 th model. However, the first set of prediction models may include only a portion of the 1 st through k-th models, and accordingly, the second set of prediction models k may include only a portion of the k +1 st through 2k-1 st models. In the case where the first set of prediction models includes the nth (2 ≦ n ≦ k) model, the second set of prediction models includes the nth + k-1 model accordingly, so that the fusion unit 106 may predict, for each portion of the resource inclusion, the nth + k-1 prediction demand for the portion in the next unit period and the nth + k + in the n-1 unit period immediately before the next unit period based on the portion obtained by the nth + k-1 modelThe ratio of the k-1 estimated demand amounts, and the nth estimated demand amount of the portion in the next unit period and n-1 unit periods immediately before the next unit period obtained by the nth model, the nth predicted demand amount of the portion in the next unit period is obtained. For example, the first set of predictive models may include only the 1 st model, the 3 rd model …, the 2j +1 st model (where,
Figure BDA0002431456350000081
) And accordingly, the second set of predictive models may include only the k +2 model … (k +2 j) th model.
As an example, the second set of predictive models may be time series models, such as generalized linear models, generalized additive models, etc., and the second set of predictive models may include a respective predictive model for each portion of the resource inclusion. In this case, the prediction result acquisition unit 104 may be further configured to generate, for each part included in the resource, a second set of prediction results regarding a required amount of the part in a next unit period after the predetermined history period, by the prediction model for the part. For example, in the example case shown in fig. 2, the second group of prediction models may include a5 th model, a6 th model, and a7 th model for each part of the resource inclusion, and the prediction result acquisition unit 104 may acquire, for each part of the resource inclusion, the 5 th to 7 th estimated demanded amounts and the 6 th and 7 th predicted demanded amounts of the part through the 5 th model, the 6 th model, and the 7 th model for the part. Further, the respective predictive models for each portion of the resource inclusion included in the second set of predictive models may be established based on historical data of the respective portions.
By way of illustration and not limitation, the first set of predictive models can be regression models, such as ridge regression models, decision tree regression models, support vector regression models, XGBoost regression models, Bagging regression models, and the like. In this case, each portion comprised by the resource may share a set of regression models. For example, in the example case shown in fig. 2, the first group of prediction models may include the 1 st, 2 nd, 3 rd and 4 th models shared by each part of the resource inclusion, and the prediction result acquisition unit 104 may obtain the 1 st predicted required amount and the 2 nd to 4 th estimated required amounts of each part of the resource inclusion through the 1 st, 2 nd, 3 rd and 4 th models.
For example, where the first set of prediction models is a regression model, the regression model may be trained by using historical data about various portions of the resource to obtain a trained regression model as the first set of prediction models.
Fig. 5 shows an example of a training process of a regression model, in which the unit period is a month. As shown in FIG. 5, historical data about various portions of a resource is first segmented by a sliding window technique to obtain training samples. For example, the demand amounts of 2018 year 11 month, 2018 year 12 month, 2019 year 1 month, and 2019 year 2 month are obtained as one training sample, where the demand amounts of 2018 year 11 month, 2018 year 12 month, 2019 year 1 month are used as X and the demand amount of 2019 year 2 month is used as Y. In addition, for example, the demand amounts of 2018 year 12 month, 2019 year 1 month, 2019 year 2 month, and 2019 year 3 month are obtained as one training sample, where the demand amounts of 2018 year 12 month, 2019 year 1 month, and 2019 year 2 month are used as X and the demand amount of 2019 year 3 month is used as Y. The regression model is then trained using the obtained training samples (where X is used as input and Y is used as a label) to obtain the trained regression model as model 1 in the first set of prediction models.
Note that, in the example shown in fig. 5, only the segmentation of the history data on one part of the resource to obtain the training samples is shown, however, in practical applications, the segmentation of the history data on the respective parts of the resource to obtain the training samples is performed in a similar manner, and the regression model is trained using all or a part of the obtained training samples to obtain the trained regression model as the 1 st model in the first set of prediction models. For example, in the case where the resource is a component of a vehicle (such as a brake pad, a tire, an oil filter, etc.), the historical data of each component of the vehicle is separately sliced in a manner similar to that shown in fig. 5 to obtain training samples, and the regression model is trained using all or a portion of the obtained training samples to obtain a trained regression model as the 1 st model in the first set of prediction models. In addition, in the example shown in fig. 5, the window size is 4, however, those skilled in the art can select an appropriate window size according to actual needs.
For the other models (i.e., 2 nd to k th models) than the 1 st model in the first set of prediction models, the corresponding processed historical data is segmented in a manner similar to that shown in fig. 4 to obtain training samples, and the regression models are trained using all or a part of the obtained training samples to obtain the trained regression models as the corresponding models in the first set of prediction models. For example, for the 2 nd model, the processed history data obtained by summing or averaging the demand amounts in the adjacent 2 unit periods is segmented to obtain training samples, and the regression model is trained using all or a part of the obtained training samples to obtain the trained regression model as the 2 nd model.
The forecast relating to the demand of the resource enables, for example, the resource provider to know in advance the number of the respective parts of the parts included in the resource which are demanded in the future, thereby preparing the respective parts included in the resource. Many techniques for predicting future demand based on historical data regarding demand have been developed. However, in practical applications, the following problems often arise: the demand in certain periods of the historical data may vary abnormally due to occasional events (e.g., promotions, weather, etc.); some models are suitable for short-term prediction, but the anti-interference capability may be insufficient; some models have strong interference rejection, but short-term accuracy may be insufficient. These problems lead to insufficient prediction accuracy and poor stability of prediction. As described above, the information processing apparatus according to the embodiment of the present disclosure obtains the final predicted demand amount in the next unit period by fusing the first group of predicted results obtained by the first group of models suitable for short-term prediction and the second group of predicted results obtained by the second group of models suitable for long-term prediction, so that the prediction accuracy and/or the antijamming capability can be improved, thereby enabling, for example, the resource preparation shortage or excess to be avoided.
In addition, the first set of predictive models may be periodically updated, thereby further improving prediction accuracy and/or interference rejection.
Corresponding to the above-described information processing apparatus embodiments, the present disclosure also provides embodiments of the following information processing method.
Fig. 6 is a flowchart illustrating an example of a flow of an information processing method 600 according to an embodiment of the present disclosure. As shown in fig. 6, the information processing method 600 according to an embodiment of the present disclosure may include a history data acquisition step S602, a prediction result acquisition step S604, and a fusion step S606.
In the history data acquisition step S602, history data about a resource may be acquired, wherein the resource contains at least one portion, and the history data includes at least a required amount of each portion contained in the resource in a plurality of unit periods respectively included in a predetermined history time period. For example, the unit period may be a season, a month, a week, or the like. For example, the history data acquiring step S602 may be implemented by the history data acquiring unit 102 described above with reference to fig. 1, and specific details are not described herein again.
In the predicted result obtaining step S604, a first set of predicted results regarding a required amount of each part included in the resource in a next unit period after the predetermined history period may be generated by a first set of prediction models; and for each portion contained by the resource, generating a second set of predictions relating to the demand for that portion for the next unit period after the predetermined historical period by a second set of prediction models. Wherein the first set of prediction models is adapted for short-term prediction and the second set of prediction models is adapted for long-term prediction. By way of illustration and not limitation, the short-term prediction accuracy of the first set of prediction models is higher than the second set of prediction models, and the long-term prediction accuracy of the second set of prediction models is higher than the first set of prediction models. For example, the prediction result obtaining step S604 may be implemented by the prediction result obtaining unit 104 described above with reference to fig. 1, and specific details are not described herein again.
In the fusing step S606, a final predicted required amount of each part included in the resource may be obtained by fusing the first group of the predicted results and the second group of the predicted results in a next unit period after the predetermined history period. For example, the fusion step S606 may be implemented by the fusion unit 106 described above with reference to fig. 1, and specific details are not described herein again.
By way of illustration and not limitation, resources may include, but are not limited to: various services, for example, regular maintenance services for vehicles such as vehicles, ships, airplanes, etc.; and various articles, for example, components included in vehicles such as vehicles, ships, airplanes, and the like; components included in electronic devices such as smart phones, portable computers, desktop computers, and the like.
As an example, the first set of predictive models may include 1 st through kth models, where k is a positive integer greater than 1, and the second set of predictive models may include k +1 st through 2k-1 st models. For example, the 1 st model included in the first set of predictive models may be obtained by training using historical data, and the ith model included in the first set of predictive models may be obtained by training using processed historical data as follows: the processed historical data is obtained by summing or averaging, for each portion of the resource containing, the demand of that portion in the historical data over i adjacent unit periods, where i is 2. The second set of predictive models may include a k +1 th model through a 2k-1 th model each established based on historical data. The first set of predictors may include: 1 st predicted demand amount in the next unit period after a predetermined history period for each part contained in the resource obtained by the 1 st model; and an ith estimated demand amount for each part contained in the resource obtained by the ith model, a next unit period after the predetermined history period, and i-1 unit periods immediately before the next unit period. The second set of predictors may include: each part contained in the resource obtained by the (i + k-1) th model included in the second group of prediction models, the (i + k-1) th predicted demand amount in the next unit period after the predetermined history period, and the (i + k-1) th estimated demand amount in the i-1 th unit period immediately before the next unit period.
In the fusion step S606, an ith predicted demand amount of the part in the next unit period may be obtained based on a ratio of an i + k-1 th predicted demand amount of the part in the next unit period obtained by the i + k-1 th model and an i + k-1 th estimated demand amount in i-1 unit periods immediately before the next unit period, and an ith estimated demand amount of the part in the next unit period and i-1 unit periods immediately before the next unit period obtained by the i model, for each part contained in the resource; and for each part contained in the resource, obtaining the final predicted demand of the part based on the 1 st predicted demand to the 2k-1 st predicted demand.
As an example, in the fusing step S606, a mean value or a median value of the 1 st to 2k-1 st predicted demand amounts may be obtained for each section included in the resource as a final predicted demand amount of the section.
As an example, the second set of predictive models may further include a 2k model built based on historical data, and the second set of predictions may further include a 2k predicted demand obtained by the 2k model. In this case, the final predicted required amount of each section may be obtained based on the 1 st to 2k th predicted required amounts of the section.
Note that, although in the above-described example of the present embodiment, the first group of prediction models includes the 1 st model to the kth model, and the second group of prediction models includes the k +1 th model to the 2k-1 th model. However, the first set of prediction models may include only a portion of the 1 st through k-th models, and accordingly, the second set of prediction models k may include only a portion of the k +1 st through 2k-1 st models. In the case where the first set of predictive models includes the nth (2. ltoreq. n.ltoreq.k) model, the second set of predictive models includes the (n + k-1) th model accordingly, so that in the fusion step S606, for each part of the resource inclusion, it is possible to base it onThe ratio of the n + k-1 th predicted demand amount of the portion in the next unit period obtained by the n + k-1 th model and the n + k-1 th estimated demand amount in the n-1 unit period immediately before the next unit period, and the n-th estimated demand amount of the portion in the next unit period and the n-1 unit period immediately before the next unit period obtained by the n-th model, the n-th predicted demand amount of the portion in the next unit period are obtained. For example, the first set of predictive models may include only the 1 st model, the 3 rd model …, the 2j +1 st model (where,
Figure BDA0002431456350000121
) And accordingly, the second set of predictive models may include only the k +2 model … (k +2 j) th model.
As an example, the second set of predictive models may be time series models, such as generalized linear models, generalized additive models, etc., and the second set of predictive models may include a respective predictive model for each portion of the resource inclusion. In this case, a second set of predicted results regarding the amount of demand of the portion in the next unit period after the predetermined history period may be generated for each portion contained in the resource by the prediction model for the portion.
By way of illustration and not limitation, the first set of predictive models can be regression models, such as ridge regression models, decision tree regression models, support vector regression models, XGBoost regression models, Bagging regression models, and the like. In this case, each portion comprised by the resource may share a set of regression models.
Note that the order of the steps of the information processing method depicted in fig. 6 is an order convenient for description, the order is not restrictive, and the actual steps are executed in parallel or in a desired order as appropriate.
The forecast relating to the demand of the resource enables, for example, the resource provider to know in advance the number of the respective parts of the parts included in the resource which are demanded in the future, thereby preparing the respective parts included in the resource. Many techniques for predicting future demand based on historical data regarding demand have been developed. However, in practical applications, the following problems often arise: the demand in certain periods of the historical data may vary abnormally due to occasional events (e.g., promotions, weather, etc.); some models are suitable for short-term prediction, but the anti-interference capability may be insufficient; some models have strong interference rejection, but short-term accuracy may be insufficient. These problems lead to insufficient prediction accuracy and poor stability of prediction. As described above, the information processing method according to the embodiment of the present disclosure obtains the final predicted demand amount in the next unit period by fusing the first group of predicted results obtained by the first group of models suitable for short-term prediction and the second group of predicted results obtained by the second group of models suitable for long-term prediction, so that the prediction accuracy and/or the antijamming capability can be improved, thereby enabling, for example, the resource preparation shortage or excess to be avoided.
In addition, the first set of predictive models may be periodically updated, thereby further improving prediction accuracy and/or interference rejection.
It should be noted that although the functional configurations and operations of the information processing apparatus and the information processing method according to the embodiments of the present disclosure are described above, this is merely an example and not a limitation, and a person skilled in the art may modify the above embodiments according to the principles of the present disclosure, for example, functional modules and operations in the respective embodiments may be added, deleted, or combined, and such modifications fall within the scope of the present disclosure.
In addition, it should be further noted that the method embodiments herein correspond to the apparatus embodiments described above, and therefore, the contents that are not described in detail in the method embodiments may refer to the descriptions of the corresponding parts in the apparatus embodiments, and the description is not repeated here.
In addition, the present disclosure also provides a storage medium and a program product. It should be understood that the machine-executable instructions in the storage medium and the program product according to the embodiments of the present disclosure may also be configured to perform the above-described information processing method, and thus, the contents not described in detail herein may refer to the description of the corresponding parts previously, and the description will not be repeated herein.
Accordingly, storage media for carrying the above-described program products comprising machine-executable instructions are also included in the present disclosure. Including, but not limited to, floppy disks, optical disks, magneto-optical disks, memory cards, memory sticks, and the like.
Further, it should be noted that the above series of processes and means may also be implemented by software and/or firmware. In the case of implementation by software and/or firmware, a program constituting the software is installed from a storage medium or a network to a computer having a dedicated hardware structure, such as a general-purpose personal computer 700 shown in fig. 7, which is capable of executing various functions and the like when various programs are installed.
In fig. 7, a Central Processing Unit (CPU)701 performs various processes in accordance with a program stored in a Read Only Memory (ROM)702 or a program loaded from a storage section 708 to a Random Access Memory (RAM) 703. In the RAM 703, data necessary when the CPU 701 executes various processes and the like is also stored as necessary.
The CPU 701, the ROM702, and the RAM 703 are connected to each other via a bus 704. An input/output interface 705 is also connected to the bus 704.
The following components are connected to the input/output interface 705: an input section 706 including a keyboard, a mouse, and the like; an output section 707 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker and the like; a storage section 708 including a hard disk and the like; and a communication section 709 including a network interface card such as a LAN card, a modem, and the like. The communication section 709 performs communication processing via a network such as the internet.
A driver 710 is also connected to the input/output interface 705 as needed. A removable medium 711 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 710 as necessary, so that the computer program read out therefrom is mounted in the storage section 708 as necessary.
In the case where the above-described series of processes is realized by software, a program constituting the software is installed from a network such as the internet or a storage medium such as the removable medium 711.
It should be understood by those skilled in the art that such a storage medium is not limited to the removable medium 711 shown in fig. 7 in which the program is stored, distributed separately from the apparatus to provide the program to the user. Examples of the removable medium 711 include a magnetic disk (including a floppy disk (registered trademark)), an optical disk (including a compact disc-read only memory (CD-ROM) and a Digital Versatile Disc (DVD)), a magneto-optical disk (including a mini-disk (MD) (registered trademark)), and a semiconductor memory. Alternatively, the storage medium may be the ROM702, a hard disk included in the storage section 708, or the like, in which programs are stored and which are distributed to users together with the apparatus including them.
The preferred embodiments of the present disclosure are described above with reference to the drawings, but the present disclosure is of course not limited to the above examples. Various changes and modifications within the scope of the appended claims may be made by those skilled in the art, and it should be understood that these changes and modifications naturally will fall within the technical scope of the present disclosure.
For example, a plurality of functions included in one unit may be implemented by separate devices in the above embodiments. Alternatively, a plurality of functions implemented by a plurality of units in the above embodiments may be implemented by separate devices, respectively. In addition, one of the above functions may be implemented by a plurality of units. Needless to say, such a configuration is included in the technical scope of the present disclosure.
In this specification, the steps described in the flowcharts include not only the processing performed in time series in the described order but also the processing performed in parallel or individually without necessarily being performed in time series. Further, even in the steps processed in time series, needless to say, the order can be changed as appropriate.
In addition, the technique according to the present disclosure can also be configured as follows.
Supplementary note 1. an information processing apparatus comprising:
a history data acquisition unit configured to acquire history data on a resource containing at least one portion, the history data including at least a required amount of each portion contained in the resource in a plurality of unit periods respectively included during a predetermined history time period;
a prediction result acquisition unit configured to:
generating, for each portion of the resource content, a first set of predictions relating to the demand of that portion for the next unit period after the predetermined historical period of time by means of a first set of prediction models; and
generating, for each portion of the resource content, a second set of prediction results relating to the demand of that portion for the next unit time period by a second set of prediction models; and
a merging unit configured to obtain, for each part included in the resource, a final predicted required amount of the part in the next unit period by merging the first group of prediction results and the second group of prediction results,
wherein the first set of prediction models is adapted for short-term prediction and the second set of prediction models is adapted for long-term prediction.
Supplementary note 2. the information processing apparatus according to supplementary note 1, wherein,
the first set of prediction models comprises a1 st model to a kth model, wherein k is a positive integer greater than 1;
the second set of predictive models includes a k +1 th model through a 2k-1 th model;
the 1 st model is obtained by training using the historical data,
the ith model is obtained by training with processed historical data as follows: the processed historical data is obtained by summing or averaging, for each portion of the resource containing, the demand of that portion in the historical data over adjacent i unit periods, where i is 2.
Each of the (k + 1) th to (2 k-1) th models is established based on the historical data;
the first set of predictors includes: 1 st predicted demand for each portion of the resource inclusion obtained by the 1 st model in the next unit period; and an ith estimated demand for each portion contained in the resource obtained by the ith model, in the next unit period and i-1 unit periods immediately before the next unit period;
the second set of predictors includes: each part contained by the resource obtained by the (i + k-1) th model included in the second set of prediction models, the (i + k-1) th predicted demand in the next unit period, and the (i + k-1) th estimated demand in the (i-1) th unit period immediately before the next unit period; and
the fusion unit is further configured to:
for each part contained in the resource, obtaining an ith predicted demand amount of the part in the next unit period based on a ratio of an i + k-1 th predicted demand amount of the part in the next unit period obtained by the i + k-1 th model and an i + k-1 th estimated demand amount in i-1 unit periods immediately before the next unit period, and an ith estimated demand amount of the part in the next unit period and i-1 unit periods immediately before the next unit period obtained by the i model; and
and for each part contained in the resources, obtaining the final predicted demand of the part based on the 1 st to 2k-1 st predicted demands.
Note 3. the information processing apparatus according to note 2, wherein the fusion unit is further configured to obtain, for each section included in the resource, a mean value or a median value of the 1 st to 2k-1 st predicted required amounts as a final predicted required amount of the section.
Note 4. the information processing apparatus according to any one of notes 1 to 3, wherein the second set of prediction models is a time series model, and the second set of prediction models includes a respective prediction model for each part included in the resource; and
wherein the prediction result acquisition unit is further configured to generate, for each part included in the resource, a second set of prediction results regarding a required amount of the part in the next unit period by the prediction model for the part.
Note 5 that the information processing apparatus according to any one of notes 1 to 3, wherein the first group of prediction models is a regression model.
Supplementary note 6. the information processing apparatus according to any one of supplementary notes 1 to 3, wherein the unit period is a season, a month, or a week.
Note 7 that the information processing apparatus according to any one of notes 1 to 3, wherein the first group of prediction models is periodically updated.
Note 8 the information processing apparatus according to note 2, wherein the second group of prediction models further includes a2 k-th model built based on the history data, and the second group of prediction results further includes each part contained in the resource obtained by the 2 k-th model, a2 k-th predicted required amount in the next unit period, and
wherein the fusion unit is further configured to obtain, for each portion included in the resource, a final predicted required amount of the portion based on the 1 st to 2k th predicted required amounts.
Supplementary note 9. an information processing method, comprising:
a history data acquisition step of acquiring history data on a resource containing at least one portion, the history data including at least a required amount of each portion contained in the resource in a plurality of unit periods respectively included in a predetermined history time period;
a prediction result acquisition step for:
generating, for each portion of the resource content, a first set of predictions relating to the demand of that portion for the next unit period after the predetermined historical period of time by means of a first set of prediction models; and
generating, for each portion of the resource content, a second set of prediction results relating to the demand of that portion for the next unit time period by a second set of prediction models; and
a merging step of, for each part included in the resource, obtaining a final predicted required amount of the part in the next unit period by merging the first group of prediction results and the second group of prediction results,
wherein the first set of prediction models is adapted for short-term prediction and the second set of prediction models is adapted for long-term prediction.
Supplementary note 10. the information processing method according to supplementary note 9, wherein,
the first set of prediction models comprises a1 st model to a kth model, wherein k is a positive integer greater than 1;
the second set of predictive models includes a k +1 th model through a 2k-1 th model;
the 1 st model is obtained by training using the historical data,
the ith model is obtained by training with processed historical data as follows: the processed historical data is obtained by summing or averaging, for each portion of the resource containing, the demand of that portion in the historical data over adjacent i unit periods, where i is 2.
Each of the (k + 1) th to (2 k-1) th models is created based on the history data;
the first set of predictors includes: 1 st predicted demand for each portion of the resource inclusion obtained by the 1 st model in the next unit period; and an ith estimated demand for each portion contained in the resource obtained by the ith model, in the next unit period and i-1 unit periods immediately before the next unit period;
the second set of predictors includes: each part contained by the resource obtained by the (i + k-1) th model included in the second set of prediction models, the (i + k-1) th predicted demand in the next unit period, and the (i + k-1) th estimated demand in the (i-1) th unit period immediately before the next unit period; and
in the fusing step:
for each part contained in the resource, obtaining an ith predicted demand amount of the part in the next unit period based on a ratio of an i + k-1 th predicted demand amount of the part in the next unit period obtained by the i + k-1 th model and an i + k-1 th estimated demand amount in i-1 unit periods immediately before the next unit period, and an ith estimated demand amount of the part in the next unit period and i-1 unit periods immediately before the next unit period obtained by the i model; and
and for each part contained in the resources, obtaining the final predicted demand of the part based on the 1 st to 2k-1 st predicted demands.
Note 11 the information processing method according to note 10, wherein in the merging step, for each part included in the resource, a mean value or a median value of the 1 st predicted required amount to the 2k-1 st predicted required amount is obtained as a final predicted required amount of the part.
Supplementary notes 12. the information processing method according to any one of supplementary notes 9 to 11, wherein the second set of prediction models is a time series model and includes a respective prediction model for each part of the resource inclusion; and
wherein, in the predicted result acquiring step, for each part included in the resource, a second group of predicted results regarding the required amount of the part in the next unit period is generated by the prediction model for the part.
Supplementary notes 13. the information processing method according to any one of supplementary notes 9 to 11, wherein the first set of prediction models is a regression model.
Supplementary notes 14. the information processing method according to any one of supplementary notes 9 to 11, wherein the unit period is a season, a month, or a week.
Supplementary note 15 the information processing method according to any one of supplementary notes 9 to 11, wherein the first group of prediction models is periodically updated.
Supplementary note 16. the information processing method according to supplementary note 10, wherein the second group of prediction models further includes a2 k-th model built based on the history data, and the second group of prediction results further includes each part contained in the resource obtained by the 2 k-th model, a2 k-th predicted required amount in the next unit period, and
wherein in the merging step, for each part included in the resource, a final predicted required amount of the part is obtained based on the 1 st to 2k th predicted required amounts.
Reference 17. a computer readable storage medium storing program instructions for performing the method of any of the references 9 to 16 when executed by a computer.

Claims (10)

1. An information processing apparatus comprising:
a history data acquisition unit configured to acquire history data on a resource containing at least one portion, the history data including at least a required amount of each portion contained in the resource in a plurality of unit periods respectively included during a predetermined history time period;
a prediction result acquisition unit configured to:
generating, for each portion of the resource content, a first set of predictions relating to the demand of that portion for the next unit period after the predetermined historical period of time by means of a first set of prediction models; and
generating, for each portion of the resource content, a second set of prediction results relating to the demand of that portion for the next unit time period by a second set of prediction models; and
a merging unit configured to obtain, for each part included in the resource, a final predicted required amount of the part in the next unit period by merging the first group of prediction results and the second group of prediction results,
wherein the first set of prediction models is adapted for short-term prediction and the second set of prediction models is adapted for long-term prediction.
2. The information processing apparatus according to claim 1,
the first set of prediction models comprises a1 st model to a kth model, wherein k is a positive integer greater than 1;
the second set of predictive models includes a k +1 th model through a 2k-1 th model;
the 1 st model is obtained by training using the historical data,
the ith model is obtained by training with processed historical data as follows: the processed historical data is obtained by summing or averaging, for each portion of the resource containing, the demand of that portion in the historical data over adjacent i unit periods, where i is 2.
Each of the (k + 1) th to (2 k-1) th models is established based on the historical data;
the first set of predictors includes: 1 st predicted demand for each portion of the resource inclusion obtained by the 1 st model in the next unit period; and an ith estimated demand for each portion contained in the resource obtained by the ith model, in the next unit period and i-1 unit periods immediately before the next unit period;
the second set of predictors includes: each part contained by the resource obtained by the (i + k-1) th model included in the second set of prediction models, the (i + k-1) th predicted demand in the next unit period, and the (i + k-1) th estimated demand in the (i-1) th unit period immediately before the next unit period; and
the fusion unit is further configured to:
for each part contained in the resource, obtaining an ith predicted demand amount of the part in the next unit period based on a ratio of an i + k-1 th predicted demand amount of the part in the next unit period obtained by the i + k-1 th model and an i + k-1 th estimated demand amount in i-1 unit periods immediately before the next unit period, and an ith estimated demand amount of the part in the next unit period and i-1 unit periods immediately before the next unit period obtained by the i model; and
and for each part contained in the resources, obtaining the final predicted demand of the part based on the 1 st to 2k-1 st predicted demands.
3. The information processing apparatus according to claim 2, wherein the fusion unit is further configured to obtain, for each section included in the resource, a mean value or a median value of the 1 st to 2k-1 st predicted required amounts as a final predicted required amount of the section.
4. The information processing apparatus according to any one of claims 1 to 3, wherein the second set of prediction models is a time series model, and the second set of prediction models includes a respective prediction model for each part of the resource inclusion; and
wherein the prediction result acquisition unit is further configured to generate, for each part included in the resource, a second set of prediction results regarding a required amount of the part in the next unit period by the prediction model for the part.
5. The information processing apparatus according to any one of claims 1 to 3, wherein the first set of prediction models is a regression model.
6. The information processing apparatus according to any one of claims 1 to 3, wherein the unit period is a season, a month, or a week.
7. The information processing apparatus according to any one of claims 1 to 3, wherein the first group of prediction models is updated periodically.
8. An information processing method comprising:
a history data acquisition step of acquiring history data on a resource containing at least one portion, the history data including at least a required amount of each portion contained in the resource in a plurality of unit periods respectively included in a predetermined history time period;
a prediction result acquisition step for:
generating, for each portion of the resource content, a first set of predictions relating to the demand of that portion for the next unit period after the predetermined historical period of time by means of a first set of prediction models; and
generating, for each portion of the resource content, a second set of prediction results relating to the demand of that portion for the next unit time period by a second set of prediction models; and
a merging step of, for each part included in the resource, obtaining a final predicted required amount of the part in the next unit period by merging the first group of prediction results and the second group of prediction results,
wherein the first set of prediction models is adapted for short-term prediction and the second set of prediction models is adapted for long-term prediction.
9. The information processing method according to claim 8,
the first set of prediction models comprises a1 st model to a kth model, wherein k is a positive integer greater than 1;
the second set of predictive models includes a k +1 th model through a 2k-1 th model;
the 1 st model is obtained by training using the historical data,
the ith model is obtained by training with processed historical data as follows: the processed historical data is obtained by summing or averaging, for each portion of the resource containing, the demand of that portion in the historical data over adjacent i unit periods, where i is 2.
Each of the (k + 1) th to (2 k-1) th models is created based on the history data;
the first set of predictors includes: 1 st predicted demand for each portion of the resource inclusion obtained by the 1 st model in the next unit period; and an ith estimated demand for each portion contained in the resource obtained by the ith model, in the next unit period and i-1 unit periods immediately before the next unit period;
the second set of predictors includes: each part contained by the resource obtained by the (i + k-1) th model included in the second set of prediction models, the (i + k-1) th predicted demand in the next unit period, and the (i + k-1) th estimated demand in the (i-1) th unit period immediately before the next unit period; and
in the fusing step:
for each part contained in the resource, obtaining an ith predicted demand amount of the part in the next unit period based on a ratio of an i + k-1 th predicted demand amount of the part in the next unit period obtained by the i + k-1 th model and an i + k-1 th estimated demand amount in i-1 unit periods immediately before the next unit period, and an ith estimated demand amount of the part in the next unit period and i-1 unit periods immediately before the next unit period obtained by the i model; and
and for each part contained in the resources, obtaining the final predicted demand of the part based on the 1 st to 2k-1 st predicted demands.
10. A computer readable storage medium storing program instructions for performing the method of claim 8 or 9 when executed by a computer.
CN202010237421.1A 2020-03-30 2020-03-30 Information processing apparatus, information processing method, and computer-readable storage medium Pending CN113469394A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010237421.1A CN113469394A (en) 2020-03-30 2020-03-30 Information processing apparatus, information processing method, and computer-readable storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010237421.1A CN113469394A (en) 2020-03-30 2020-03-30 Information processing apparatus, information processing method, and computer-readable storage medium

Publications (1)

Publication Number Publication Date
CN113469394A true CN113469394A (en) 2021-10-01

Family

ID=77865036

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010237421.1A Pending CN113469394A (en) 2020-03-30 2020-03-30 Information processing apparatus, information processing method, and computer-readable storage medium

Country Status (1)

Country Link
CN (1) CN113469394A (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20120036568A (en) * 2010-10-08 2012-04-18 주식회사 케이디파워 Method for prediciting power load and method for sampling pattern
JP2013131259A (en) * 2013-04-04 2013-07-04 Ntt Data Corp Integrated demand prediction device, integrated demand prediction method, and integrated demand prediction program
US20140371934A1 (en) * 2012-02-14 2014-12-18 Toshiba Mitsubishi-Electric Industrial Systems Corporation Electricity demand prediction system
CN108805313A (en) * 2017-04-26 2018-11-13 株式会社日立制作所 Needing forecasting method, device towards global supply chain and system
CN110033292A (en) * 2018-01-12 2019-07-19 北京京东尚科信息技术有限公司 Information output method and device
CN110390425A (en) * 2019-06-20 2019-10-29 阿里巴巴集团控股有限公司 Prediction technique and device
CN110555537A (en) * 2018-05-30 2019-12-10 微软技术许可有限责任公司 Multi-factor multi-time point correlated prediction

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20120036568A (en) * 2010-10-08 2012-04-18 주식회사 케이디파워 Method for prediciting power load and method for sampling pattern
US20140371934A1 (en) * 2012-02-14 2014-12-18 Toshiba Mitsubishi-Electric Industrial Systems Corporation Electricity demand prediction system
JP2013131259A (en) * 2013-04-04 2013-07-04 Ntt Data Corp Integrated demand prediction device, integrated demand prediction method, and integrated demand prediction program
CN108805313A (en) * 2017-04-26 2018-11-13 株式会社日立制作所 Needing forecasting method, device towards global supply chain and system
CN110033292A (en) * 2018-01-12 2019-07-19 北京京东尚科信息技术有限公司 Information output method and device
CN110555537A (en) * 2018-05-30 2019-12-10 微软技术许可有限责任公司 Multi-factor multi-time point correlated prediction
CN110390425A (en) * 2019-06-20 2019-10-29 阿里巴巴集团控股有限公司 Prediction technique and device

Similar Documents

Publication Publication Date Title
CN102687166B (en) For the method and apparatus of user interest modeling
US9430299B2 (en) System, method and product for task allocation
US20160148251A1 (en) Risk Quantification for Policy Deployment
US11676060B2 (en) Digital content interaction prediction and training that addresses imbalanced classes
CN110298716B (en) Information pushing method and device
CN111125574A (en) Method and apparatus for generating information
US11676075B2 (en) Label reduction in maintaining test sets
CN112508118A (en) Target object behavior prediction method aiming at data migration and related equipment thereof
CN111027707A (en) Model optimization method and device and electronic equipment
CN109389424B (en) Flow distribution method and device, electronic equipment and storage medium
CN112989146A (en) Method, apparatus, device, medium, and program product for recommending resources to a target user
CN115801980A (en) Video generation method and device
CN114139052B (en) Ranking model training method for intelligent recommendation, intelligent recommendation method and device
US20160148246A1 (en) Automated System for Safe Policy Improvement
CN113159453B (en) Resource data prediction method, device, equipment and storage medium
CN115018552A (en) Method for determining click rate of product
Liu et al. Cliqueparcel: An approach for batching llm prompts that jointly optimizes efficiency and faithfulness
CN113297478B (en) Information pushing method and device based on user life cycle and electronic equipment
CN113469394A (en) Information processing apparatus, information processing method, and computer-readable storage medium
CN109951859B (en) Wireless network connection recommendation method and device, electronic equipment and readable medium
CN114742645B (en) User security level identification method and device based on multi-stage time sequence multitask
US20190114673A1 (en) Digital experience targeting using bayesian approach
CN112070530A (en) Online evaluation method and related device of advertisement prediction model
CN113360773A (en) Recommendation method and device, storage medium and electronic equipment
CN111339432A (en) Recommendation method and device of electronic object and electronic equipment

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