CN112132323A

CN112132323A - Method and device for predicting value amount of commodity object and electronic equipment

Info

Publication number: CN112132323A
Application number: CN202010866132.8A
Authority: CN
Inventors: 裴玉; 马俊; 范勐喆; 崔一诺; 舒平平; 马红
Original assignee: Hanhai Information Technology Shanghai Co Ltd
Current assignee: Hanhai Information Technology Shanghai Co Ltd
Priority date: 2020-08-25
Filing date: 2020-08-25
Publication date: 2020-12-25

Abstract

The disclosure provides a method, a device and an electronic device for predicting value quantity of a commodity object, wherein the method comprises the following steps: acquiring the actual value quantity of the commodity object in the historical use period in the corresponding historical purchase period; determining the initial predicted value amount of the commodity object in the target purchase period in the target use period according to the historical use period and the corresponding actual value amount; acquiring a target vector value of a characteristic vector of a commodity object in a target use period in a target purchase period and a mapping function between the characteristic vector and a fluctuation score; according to the target vector value and the mapping function, determining the predicted rising and falling scores of the commodity objects in the target use period in the target purchase period; and obtaining the target predicted value quantity of the commodity object in the target use period in the target purchase period according to the initial predicted value quantity and the predicted rising and falling scores.

Description

Method and device for predicting value amount of commodity object and electronic equipment

Technical Field

The present disclosure relates to the field of model prediction technologies, and in particular, to a method for predicting a value amount of a commodity object, an apparatus for predicting a value amount of a commodity object, and an electronic device.

Background

An airplane is one of the important vehicles for long distance travel of users. Under the influence of factors such as airline policies and supply and demand relations, the air tickets usually show price fluctuation of different degrees at different time points before taking off, so that users usually pay attention to and track the prices of the air tickets before buying the air tickets, and the travel dates or the ticket buying times are selected according to the actual demands of the users and the prices of the air tickets.

Under a common condition, a user is difficult to accurately judge the change rule and the change amplitude of the air ticket price according to own experience, so that the user is difficult to purchase the air ticket which not only meets the self travel requirement, but also meets the self budget in advance, and the ticket purchasing experience of the user is influenced.

Disclosure of Invention

It is an object of the present disclosure to provide a new technical solution for predicting the value amount of a commodity object.

According to a first aspect of the present disclosure, there is provided a method for predicting a value amount of a commodity object, comprising:

acquiring the actual value quantity of the commodity object in the historical use period in the corresponding historical purchase period;

determining an initial predicted value amount of the commodity object in a target purchase period in the target use period according to the historical use period and the corresponding actual value amount; the first period time length is the same as the second period time length, the first period time length is the period time length which is the difference between the historical using period and the corresponding historical purchasing period, and the second period time length is the period time length which is the difference between the target using period and the target ticket purchasing period;

acquiring a target vector value of a feature vector of the commodity object in the target use period in the target purchase period and a mapping function between the feature vector and a rising and falling score; wherein the feature vector comprises a plurality of features that influence the rise and fall tendency of the merchandise object in the corresponding purchase period; the rising and falling scores are scores representing rising and falling trends of the commodity objects in corresponding purchase periods;

according to the target vector value and the mapping function, determining a predicted rising and falling score of the commodity object in the target use period in the target purchase period;

and obtaining the target predicted value quantity of the commodity object in the target purchase period in the target use period according to the initial predicted value quantity and the predicted rising and falling scores.

Optionally, the determining an initial predicted value amount of the commodity object in the target purchase cycle in the target use cycle according to the historical use cycle and the corresponding actual value amount of the historical purchase cycle includes:

constructing a prediction time sequence according to the historical service cycle and the corresponding actual value quantity;

and determining the initial predicted value amount of the commodity object in the target purchase period in the target use period according to the predicted time sequence based on a preset time sequence prediction model.

Optionally, the method further includes:

under the condition that two adjacent historical use periods are not continuous, determining the missing historical use period between the two historical use periods according to the period duration of the historical use period and the period duration of the difference between the two historical use periods;

and filling up the missing historical use period between the two historical use periods, and marking the actual value quantity of the commodity object in the filled historical use period in the corresponding historical purchase period as a set value.

Optionally, the method further includes:

determining the effective range of the actual value quantity of the commodity objects in the historical use period in the corresponding historical purchase period;

and eliminating actual value quantity exceeding the effective range and corresponding historical use period.

Optionally, the determining, based on a preset time series prediction model and according to the predicted time series, the initial predicted value amount of the commodity object in the target purchase period in the target usage period includes:

inputting the predicted time sequence into the time sequence prediction model to obtain a prediction result of the value amount of the commodity object in the target purchase period in the target use period;

determining at least one reference value according to the actual value quantity;

and selecting a reference value with the minimum distance from the prediction result as the initial prediction value quantity.

Optionally, the determining at least one reference value according to the actual value quantity includes:

clustering the actual value quantity based on a preset classification distance to obtain at least one value quantity classification;

and selecting the clustering center corresponding to the value quantity classification from the actual value quantities contained in the value quantity classification as the reference value.

Optionally, the method further includes:

acquiring actual value quantities of the commodity objects in a plurality of training use periods in corresponding training purchase periods; the third period time length is the same as the first period time length, and the third period time length is the period time length of the difference between the training use period and the corresponding training purchase period;

constructing a first training sample according to the training use period and the actual value quantity of the commodity object in the training use period in the corresponding training purchase period; wherein the first training sample comprises a training time series and an actual value quantity which are matched;

and performing machine learning training according to the first training sample based on a preset Prophet algorithm to obtain the time series prediction model.

Optionally, the performing machine learning training according to the first training sample based on a preset Prophet algorithm to obtain the time series prediction model includes:

determining a predicted value amount expression of the first training sample by taking a pending parameter of the Prophet algorithm as a variable according to the training time sequence of the first training sample;

determining an average absolute percentage error between the predicted value expression of the first training sample and the actual value corresponding to the first training sample as the objective function;

and determining the undetermined parameters according to the loss function, and finishing the training of the time series prediction model.

Optionally, obtaining the target predicted value amount of the commodity object in the target purchase period according to the initial predicted value amount and the predicted rise-fall score includes:

according to the value weight corresponding to the initial prediction value and the fraction weight corresponding to the prediction rising and falling fraction;

and carrying out weighted summation on the initial predicted value quantity and the predicted rising and falling scores according to the value quantity weight and the score weight to obtain the target predicted value quantity of the commodity object in the target purchase period in the target use period.

Optionally, the step of obtaining a mapping function between the feature vector and the fluctuation fraction includes:

acquiring the actual value quantity of the commodity object in the training use period in the corresponding training purchase period and the actual value quantity in the corresponding next purchase period; wherein the next purchase cycle is a next cycle of the training purchase cycle;

comparing the actual value amount of the next purchase period with the actual value amount of the training purchase period, and determining the actual rising and falling score of the commodity object corresponding to the training use period according to the comparison result;

acquiring a training vector value of the feature vector of the commodity object in the training use period in the training purchase period;

generating a second training sample according to the training vector value and the actual rise-fall score;

and training to obtain the mapping function according to the training vector value of the feature vector of the second training sample and the actual rise-fall score corresponding to the second training sample.

Optionally, the method further includes:

and displaying the target predicted value amount of the commodity object of the target use period in the target purchase period, so that a user can select the purchase period of the commodity object of the target use period according to the target predicted value amount.

According to a second aspect of the present disclosure, there is provided an apparatus for predicting a value amount of a commodity object, including:

the first acquisition module is used for acquiring the actual value quantity of the commodity object in the historical use period in the corresponding historical purchase period;

the first prediction module is used for determining the initial predicted value amount of the commodity object in the target purchase period in the target use period according to the historical use period and the corresponding actual value amount; the first period time length is the same as the second period time length, the first period time length is the period time length which is the difference between the historical using period and the corresponding historical purchasing period, and the second period time length is the period time length which is the difference between the target using period and the target ticket purchasing period;

the second acquisition module is used for acquiring a target vector value of the characteristic vector of the commodity object in the target use period in the target purchase period and a mapping function between the characteristic vector and the rising and falling scores; wherein the feature vector comprises a plurality of features that influence the rise and fall tendency of the merchandise object in the corresponding purchase period; the rising and falling scores are scores representing rising and falling trends of the commodity objects in corresponding purchase periods;

the second prediction module is used for determining the predicted rising and falling scores of the commodity objects in the target use period in the target purchase period according to the target vector value and the mapping function;

and the third prediction module is used for obtaining the target predicted value quantity of the commodity object in the target purchase period in the target use period according to the initial predicted value quantity and the predicted rising and falling scores.

According to a third aspect of the present disclosure, there is provided an electronic device comprising:

the apparatus according to the second aspect of the present disclosure, or,

a processor and a memory for storing an executable computer program for controlling the processor to perform the method according to the first aspect of the present disclosure.

One beneficial effect of the present disclosure is that, by the embodiments of the present disclosure, the initial predicted value amount of the commodity object in the target purchase cycle of the target usage cycle is determined according to the historical usage cycle and the actual value amount of the commodity object in the historical purchase cycle corresponding to the historical usage cycle; determining the predicted rising and falling scores of the commodity objects in the target use period in the target purchase period according to the target vector value of the commodity objects in the target use period in the feature vector of the target purchase period and the mapping function between the feature vector and the rising and falling scores; according to the initial predicted value amount and the predicted rising and falling scores, the target predicted value amount of the commodity object in the target use period in the target purchase period is determined, so that the accuracy of the finally obtained target predicted value amount is higher, and further, accurate guidance can be provided for the behavior of purchasing the commodity object by the user.

Like this, the user can select the purchase cycle and the life cycle of commodity object according to self demand for the user can also enjoy comparatively preferential price when satisfying self trip demand, promotes user's purchase experience.

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

The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the disclosure and together with the description, serve to explain the principles of the disclosure.

Fig. 1 is a block diagram of one example of a hardware configuration of an electronic device that may be used to implement embodiments of the present disclosure.

Fig. 2 is a block diagram of another example of a hardware configuration of an electronic device that may be used to implement embodiments of the present disclosure.

Fig. 3 is a flowchart illustrating a method of predicting a value amount of a commodity object according to an embodiment of the present disclosure.

Fig. 4 is a schematic block diagram of a value amount prediction apparatus for a commodity object according to an embodiment of the present disclosure.

FIG. 5 illustrates a functional block diagram of an electronic device of one embodiment of the present disclosure.

Detailed Description

Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that: the relative arrangement of the components and steps, the numerical expressions, and numerical values set forth in these embodiments do not limit the scope of the present disclosure unless specifically stated otherwise.

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.

Techniques, methods, and apparatus known to those of ordinary skill in the relevant art may not be discussed in detail but are intended to be part of the specification where appropriate.

In all examples shown and discussed herein, any particular value should be construed as merely illustrative, and not limiting. Thus, other examples of the exemplary embodiments may have different values.

It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, further discussion thereof is not required in subsequent figures.

< hardware configuration >

Fig. 1 and 2 are block diagrams of hardware configurations of electronic devices that can be used to implement embodiments of the present disclosure.

In one embodiment, as shown in FIG. 1, the electronic device 1000 may be a server 1100.

The server 1100 is a service point that provides processing, databases, and communications facilities. The server 1100 can be a unitary server or a distributed server across multiple computers or computer data centers. The server may be of various types, such as, but not limited to, a web server, a news server, a mail server, a message server, an advertisement server, a file server, an application server, an interaction server, a database server, or a proxy server. In some embodiments, each server may include hardware, software, or embedded logic components or a combination of two or more such components for performing the appropriate functions supported or implemented by the server.

In an embodiment, the server may be a blade server, a rack server, or a cloud server, or may be a server group composed of a plurality of servers, or may be implemented as a cloud architecture, for example, implemented by a server cluster deployed in a cloud, and may further include one or more of the above types of servers.

In this embodiment, the server 1100 may include a processor 1110, a memory 1120, an interface device 1130, a communication device 1140, a display device 1150, and an input device 1160, as shown in fig. 1.

In this embodiment, the server 1100 may also include a speaker, a microphone, and the like, which are not limited herein.

The processor 1110 may be a dedicated server processor, or may be a desktop processor, a mobile version processor, or the like that meets performance requirements, and is not limited herein. The memory 1120 includes, for example, a ROM (read only memory), a RAM (random access memory), a nonvolatile memory such as a hard disk, and the like. The interface device 1130 includes various bus interfaces such as a serial bus interface (including a USB interface), a parallel bus interface, and the like. The communication device 1140 is capable of wired or wireless communication, for example. The display device 1150 is, for example, a liquid crystal display panel, an LED display panel touch display panel, or the like. Input devices 1160 may include, for example, a touch screen, a keyboard, and the like.

In this embodiment, the memory 1120 of the server 1100 is configured to store instructions for controlling the processor 1110 to operate at least to perform a method of predicting a value amount of a commodity object according to any of the embodiments of the present disclosure. The skilled person can design the instructions according to the disclosed solution of the present disclosure. How the instructions control the operation of the processor is well known in the art and will not be described in detail herein.

Although a number of devices of server 1100 are shown in fig. 1, the present disclosure may refer to only some of the devices, e.g., server 1100 refers to only memory 1120 and processor 1110.

In one embodiment, the electronic device 1000 may be a terminal device 1200 such as a PC, a notebook computer, etc. used by an operator, as shown in fig. 2, which is not limited herein.

In this embodiment, referring to fig. 2, the terminal apparatus 1200 may include a processor 1210, a memory 1220, an interface device 1230, a communication device 1240, a display device 1250, an input device 1260, a speaker 1270, a microphone 1280, and the like.

The processor 1210 may be a mobile version processor. The memory 1220 includes, for example, a ROM (read only memory), a RAM (random access memory), a nonvolatile memory such as a hard disk, and the like. The interface device 1230 includes, for example, a USB interface, a headphone interface, and the like. The communication device 1240 may be capable of wired or wireless communication, for example, the communication device 1240 may include a short-range communication device, such as any device that performs short-range wireless communication based on short-range wireless communication protocols, such as the Hilink protocol, WiFi (IEEE 802.11 protocol), Mesh, bluetooth, ZigBee, Thread, Z-Wave, NFC, UWB, LiFi, and the like, and the communication device 1240 may also include a long-range communication device, such as any device that performs WLAN, GPRS, 2G/3G/4G/5G long-range communication. The display device 1250 is, for example, a liquid crystal display, a touch display, or the like. The input device 1260 may include, for example, a touch screen, a keyboard, and the like. A user can input/output voice information through the speaker 1270 and the microphone 1280.

In this embodiment, the memory 1220 of the terminal device 1200 is configured to store instructions for controlling the processor 1210 to operate at least to perform a method of predicting a value amount of a commodity object according to any of the embodiments of the present disclosure. The skilled person can design the instructions according to the disclosed solution of the present disclosure. How the instructions control the operation of the processor is well known in the art and will not be described in detail herein.

Although a plurality of devices of the terminal apparatus 1200 are shown in fig. 2, the present disclosure may refer to only some of the devices, for example, the terminal apparatus 1200 refers to only the memory 1220 and the processor 1210 and the display device 1250.

According to the historical use period and the actual value quantity of the commodity object in the historical use period in the corresponding historical purchase period, determining the initial predicted value quantity of the commodity object in the target purchase period in the target use period; determining the predicted rising and falling scores of the commodity objects in the target use period in the target purchase period according to the target vector value of the commodity objects in the target use period in the feature vector of the target purchase period and the mapping function between the feature vector and the rising and falling scores; according to the initial predicted value amount and the predicted rising and falling scores, the target predicted value amount of the commodity object in the target use period in the target purchase period is determined, so that the accuracy of the finally obtained target predicted value amount is higher, and further, accurate guidance can be provided for the behavior of purchasing the commodity object by the user. Like this, the user can select the purchase cycle and the life cycle of commodity object according to self demand for the user can also enjoy comparatively preferential price when satisfying self trip demand, promotes user's purchase experience.

< method examples >

The present disclosure provides a method of predicting a value amount of a commodity object. The method can be implemented by the server or the terminal device alone, or can be implemented by the server and the terminal device together. In one embodiment, the terminal device may be the terminal device 1200 as shown in FIG. 2 and the server may be the server 1100 as shown in FIG. 1.

In embodiments of the present disclosure, the value amount of the merchandise object fluctuates in different purchase cycles prior to use. The amount of value in this disclosure may be a selling price. The merchandise object may be, for example, a ticket specifying a flight, a ticket specifying a travel party, a lodging certificate specifying a hotel, or a movie ticket specifying a session, etc.

When the commodity object may be a ticket of a designated flight, for example, the statistical cycle corresponding to the departure time of the designated flight may be a usage cycle of the ticket of the designated flight; before the designated flight takes off, the statistical period corresponding to the time when the price of the ticket of the designated flight is checked can be a purchase period of the ticket of the designated flight.

Table 1 is a value volume data table for a given flight. In the example shown in table 1 below, the cycle duration of the use cycle and the purchase cycle may both be 1 day.

TABLE 1

In the case that the departure date of the designated flight is 2019-06-04, the use period of the ticket of the designated flight can be 2019-06-04, and the purchase period of the ticket of the designated flight can include the day of departure (2019-06-04), 1 day before departure (2019-06-03), 2 days before departure (2019-06-02), 3 days before departure (2019-06-01), 4 days before departure (2019-05-31), … …, and n days before departure.

In the case that the takeoff date of the designated flight is 2019-06-10, the use period of the ticket of the designated flight can be 2019-06-10, and the purchase period of the ticket of the designated flight can include the day of takeoff (2019-06-10), 1 day before takeoff (2019-06-09), 2 days before takeoff (2019-06-08), 3 days before takeoff (2019-06-07), 4 days before takeoff (2019-06-06), … …, and n days before takeoff.

In the embodiment, the price of the ticket of the designated flight with the takeoff date of 2019-06-16 in the corresponding 1 day (2019-06-15) before the takeoff can be predicted according to the price of the ticket of the designated flight with the takeoff date of 2019-06-03-2019-06-15 in the corresponding 1 day before the takeoff. Wherein the price of the ticket of the designated flight with the take-off date of 2019-06-03, which is 1 day (2019-06-02) before the take-off, can be 540; the price 1 day before takeoff (2019-06-03) corresponding to the ticket of the designated flight with the takeoff date of 2019-06-04 can be 1270; the price of the ticket of the designated flight with the take-off date of 2019-06-05, which corresponds to 1 day (2019-06-04) before the take-off, can be 1260; the price 1 day before takeoff (2019-06-05) corresponding to the ticket of the designated flight with the takeoff date of 2019-06-06 can be 1480; the price 1 day before takeoff (2019-06-06) corresponding to the ticket of the designated flight with the takeoff date of 2019-06-07 can be 600; the price 1 day before takeoff (2019-06-07) corresponding to the ticket for the designated flight with the takeoff date of 2019-06-08 may be 400; the price 1 day before takeoff (2019-06-08) corresponding to the ticket of the designated flight with the takeoff date of 2019-06-09 can be 1090; the price 1 day before takeoff (2019-06-09) corresponding to the ticket of the designated flight with the takeoff date of 2019-06-10 may be 820; the price of 1 day (2019-06-10) before the take-off corresponding to the ticket of the designated flight with the take-off date of 2019-06-11 can be 0; the price 1 day before takeoff (2019-06-11) corresponding to the ticket for the designated flight with the takeoff date of 2019-06-12 may be 1250; the price of the ticket of the designated flight with the take-off date of 2019-06-13, which corresponds to 1 day (2019-06-12) before the take-off, can be 1260; the price 1 day before takeoff (2019-06-13) corresponding to the ticket of the designated flight with the takeoff date of 2019-06-14 may be 1400; the ticket for a given flight with a takeoff date of 2019-06-15 may correspond to a price of 450 days (2019-06-14) from takeoff.

FIG. 3 is a schematic flow chart diagram of a method of predicting an amount of value of a commodity object according to an embodiment of the present disclosure.

As shown in fig. 3, the method of the present embodiment includes the following steps S3100 to S3400:

in step S3100, the actual value amount of the commodity object in the history use period in the corresponding history purchase period is acquired.

In one embodiment of the present disclosure, the historical usage period and the corresponding historical purchase period may be determined in advance according to an object to be predicted (a target predicted value amount of the commodity object of the target usage period in the target purchase period).

The historical usage period is a statistical period prior to the target usage period. The cycle duration of any use cycle and any purchase cycle is the same, and the cycle duration can be preset according to application scenes or specific requirements. For example, the period duration may be 1 day.

The first period time length is the same as the second period time length, the first period time length is the period time length of the difference between the historical using period and the corresponding historical purchasing period, and the second period time length is the period time length of the difference between the target using period and the target purchasing period.

For example, in the embodiment shown in Table 1, if the target usage period is 2019-06-16 and the target purchase period is 2019-06-14, the difference between the target usage period and the target purchase period may be 2 periods long, i.e., 2 days. Then, the value amount of the commodity object of the target use period in the target purchase period can be the purchase price of the airplane ticket with the takeoff date of 2019-06-16 2 days before takeoff (2019-06-14).

Correspondingly, the historical usage period may be a set number of statistical periods prior to 2019-06-15 (including 2019-06-15). The set number may be set in advance according to an application scenario or a specific requirement. For example, in the case where one of the historical usage periods is 2019-06-15, the corresponding historical purchase period may be 2019-06-13; in the case where one of the historical usage periods is 2019-06-14, the corresponding historical purchase period may be 2019-06-12; in the case where one of the historical usage periods is 2019-06-13, the corresponding historical purchase period may be 2019-06-11; in the case where one of the historical usage periods is 2019-06-12, the corresponding historical purchase period may be 2019-06-10; in the case where one of the historical usage periods is 2019-06-11, the corresponding historical purchase period may be 2019-06-09; in the case where one of the historical usage periods is 2019-06-10, the corresponding historical purchase period may be 2019-06-08; in the case where one of the historical usage periods is 2019-06-09, the corresponding historical purchase period may be 2019-06-07; in the case where one of the historical usage periods is 2019-06-08, the corresponding historical purchase period may be 2019-06-06; in the case where one of the historical usage periods is 2019-06-07, the corresponding historical purchase period may be 2019-06-05; in the case where one of the historical usage periods is 2019-06-06, the corresponding historical purchase period may be 2019-06-04; in the case where one of the historical usage periods is 2019-06-05, the corresponding historical purchase period may be 2019-06-03; in the case where one of the historical use periods is 2019-06-04, the corresponding historical purchase period may be 2019-06-02; in the case where one of the historical usage periods is 2019-06-03, the corresponding historical purchase period may be 2019-06-01.

Step S3200, according to the historical use cycle and the actual value quantity, determining the initial predicted value quantity of the commodity object in the target purchase cycle in the target use cycle.

In one embodiment of the present disclosure, determining the initial predicted value amount of the commodity object in the target purchase period according to the historical usage period and the actual value amount may include steps S3210 to S3220 as follows:

step S3210, a prediction time sequence is constructed according to the historical use period and the corresponding actual value amount.

The prediction time sequence of this embodiment may include: the commodity objects in the plurality of historical use periods and the actual value amount of each commodity object in each historical use period in the corresponding historical purchase period are arranged according to the time sequence.

In one embodiment of the present disclosure, the historical usage periods in the predicted time series need to be continuous. Then, before performing the step S3210, the method may further include: when two adjacent history use periods are not continuous, determining the missing history use period between the two history use periods according to the period duration of the history use period and the period duration of the difference between the two history use periods; and filling up the missing historical use period between the two historical use periods, and marking the actual value quantity of the commodity object with the filled historical use period in the corresponding historical purchase period as a set value.

In this embodiment, the number of the missing history use cycles may be specifically determined according to the cycle duration of the history use cycles and the cycle duration of the difference between the two history use cycles; and determining the missing historical use period between the two historical use periods according to the number of the missing historical use periods and the period duration of the historical use period.

The cycle duration of the missing historical use cycle is the same as the cycle duration of any historical use cycle, and the filled historical use cycle can be continuous with any one of two discontinuous adjacent historical use cycles by filling the missing historical use cycle.

For example, in the case that the duration of the history use period is 1 day, and two discontinuous adjacent history use periods are 2019-05-10 and 2019-05-15, respectively, the period duration of the period between the two history use periods is 4 days, then the history use periods missing between the two history use periods may include: 2019-05-11, 2019-05-12, 2019-05-13 and 2019-05-14.

For another example, in a case that the duration of the history use period is 1 hour, and two discontinuous adjacent history use periods are respectively 9 to 10 points and 13 to 14 points, the period duration of the difference between the two history use periods is 3 hours, then the history use period missing between the two history use periods may include: 10 to 11 points, 11 to 12 points and 12 to 13 points.

In one embodiment of the present disclosure, the setting value may be set in advance according to an application scenario or a specific requirement. For example, the set value may be, but is not limited to, 0.

In the embodiment, the missing historical use periods are filled, and the actual value quantities of the commodity objects in the filled historical use periods in the corresponding historical purchase periods are marked as the set values, so that the filled multiple historical use periods are continuous to generate the prediction time sequence for the subsequent generation, and the initial prediction value quantities of the commodity objects in the target use periods in the target purchase periods are convenient to predict.

In this embodiment, two adjacent historical periods of use are not consecutive, which may indicate that the commodity object with the missing historical period of use cannot be purchased in the corresponding historical period of use, specifically, the commodity object with the missing historical period of use is sold out in the corresponding historical period of use, or the commodity object cannot be used in the missing historical period of use.

In an embodiment where the commodity object is a ticket for a designated flight, the actual value amount of the commodity object of the historical usage period acquired through the above step S3100 in the corresponding historical purchase period may be as shown in table 2 below. If the designated flight has no flight plan at 2019-06-11 or the ticket for the designated flight has sold out before 2019-06-11, the departure date for the designated flight may be discontinuous. Then, under the condition that the price of the ticket of the designated flight with the take-off time of 2019-06-09-2019-06-15 in 2 days before the take-off is predicted according to the price of the ticket of the designated flight with the take-off time of 2019-06-16 in 2 days before the take-off, two adjacent historical use periods 2019-06-10 and 2019-06-12 are not continuous.

TABLE 2

Takeoff date (year-month-day)	Price 2 days before takeoff
		2019-06-09	1020
2019-06-10	660
		2019-06-12	1130
2019-06-13	1120
		2019-06-14	1400
2019-06-15	450

According to the cycle duration (1 day) of the historical use cycle, the missing takeoff time between two adjacent takeoff times of 2019-06-10 and 2019-06-12 can be determined to be 2019-06-11. In order to construct a predicted time sequence with continuous takeoff time, the missing takeoff time 2019-06-11 can be filled between two adjacent takeoff times 2019-06-10 and 2019-06-12, the price of the ticket of the designated flight with the filled takeoff time 2019-06-11, which is 2 days before takeoff, is marked as a set value 0, and the predicted time sequence obtained after filling can be as shown in the following table 3.

TABLE 3

2019-06-09	1020
		2019-06-10	660
2019-06-11	0
		2019-06-12	1130
2019-06-13	1120
		2019-06-14	1400
2019-06-15	450

In an embodiment of the present disclosure, before performing the step S3210, the method may further include:

determining the effective range of the actual value quantity of the commodity objects in the historical use period in the corresponding historical purchase period; and eliminating actual value quantity beyond the effective range and the corresponding historical use period.

In one example, the value of the commodity object in the historical use period is determined as an average value of the actual value of the commodity object in the corresponding historical purchase period, and the effective range is determined according to the average value and a preset error range. The error range may be a percentage or a value amount preset according to an application scenario or a specific requirement.

For example, the error range may be a percentage, and then, in the case where the average value is a and the error range is ± 50%, the effective range may be [ -a × 50%, a × 50% ].

As another example, the error range may be a value amount, and then, in the case where the average value is a and the error range is + -b, the valid range may be [ a-b, a + b ].

In another example, a median of the actual value amounts of the commodity objects in the historical use period in the corresponding historical purchase period may be determined, and the valid range may be determined according to the median and a preset error range.

The method for determining the effective range according to the median and the error range may be the same as the method for determining the effective range according to the average and the error range, and specific reference may be made to the foregoing embodiments, which are not described herein again.

In an embodiment of the present disclosure, if the historical usage period and the corresponding actual value amount are filled in advance, the effective range may not be determined according to the actual value amount corresponding to the filled historical usage period, or the actual value amount corresponding to the filled historical usage period may not be removed.

In this embodiment, the actual value amount exceeding the effective range may have a low reference value to the value amount of the commodity object in the target purchase period in the target use period, and may interfere with the prediction result, so that the accuracy of predicting the value amount of the commodity object in the target purchase period in the target use period may be improved by eliminating the actual value amount exceeding the effective range and the corresponding historical use period.

Step S3220, based on the preset time series prediction model, according to the prediction time series, determining an initial predicted value amount of the commodity object in the target purchase period in the target usage period.

The time series prediction model in this embodiment may be trained in advance, and a corresponding prediction result may be obtained according to the input time series.

Based on the time series prediction model, according to the prediction time series, the value quantity of the commodity object in the target use period in the target purchase period can be accurately predicted according to the change rule of the actual value quantity of the commodity object in the historical use period in the corresponding historical purchase period.

In one embodiment of the disclosure, the output result of the time-series prediction model may only include the prediction result of the value amount of the commodity object in the target usage period in the target purchase period, and may further include the prediction result of the value amount of the commodity object in the target usage period in the target purchase period, and the prediction results of the value amounts of the commodity objects in other usage periods in the corresponding purchase period, where the other usage periods are usage periods after the historical usage period.

In one embodiment of the present disclosure, the time series of predictions may be input into the time series prediction model, and the result of predicting the value amount of the commodity object in the target purchase cycle in the target usage cycle, which is output by the time series prediction model, may be used as the initial predicted value amount.

In another embodiment of the present disclosure, the prediction results output by the time series prediction model are continuous values, but the value amount of the commodity object may be discrete. Therefore, in order to make the obtained initial value amount more accurate, determining the initial predicted value amount of the commodity object in the target purchase period according to the predicted time series based on the preset time series prediction model may further include steps S3221 to S3223 as follows:

step S3221 is to input the prediction time series into the time series prediction model, and obtain a prediction result of the value amount of the commodity object in the target purchase period in the target usage period.

Step S3222 determines at least one reference value according to the actual value amount.

In one embodiment of the present disclosure, the value of the actual value amount may be taken as a reference value.

For example, in the embodiment shown in table 2 above, the reference values may include 1090, 820, 1250, 1260 and 450.

In another embodiment of the present disclosure, determining at least one reference value from the actual value quantity may include:

clustering the actual value quantity based on a preset classification distance to obtain at least one value quantity classification; and selecting the clustering center corresponding to the value quantity classification from the actual value quantity contained in at least one value quantity classification as a reference value.

The classification distance may be set in advance according to an application scenario or a specific requirement.

Each value amount classification obtained by clustering the actual value amounts based on the preset classification distance may include at least one actual value amount. For each value measure classification, a corresponding cluster center may be selected from the at least one actual value measure included as a reference value. I.e. the cluster center may be one of the actual value quantities contained in the corresponding value quantity classification.

For example, one value amount classification may include actual value amounts 1250, 1260, and 1260 as follows, and then actual value amount 1260 may be selected as a cluster center of the value amount classification and the cluster center may be used as a reference value.

Step S3223, a reference value having the smallest distance to the prediction result is selected as the initial prediction value.

Specifically, the distance between the prediction result obtained in step S3221 and each reference value may be selected, and the reference value having the smallest distance from the prediction result may be selected as the initial prediction value amount.

The distance in this embodiment may be, for example, a euclidean distance, or an absolute value of the difference.

In the general case, the value volume of a commodity object may be approximated by finding a corresponding mapping in its historical value volume. Therefore, the method of the embodiment discretizes the prediction result through the clustering algorithm, so that the deviation between the initial predicted value quantity and the actual value quantity of the commodity object in the target use period in the target purchase period can be reduced, and the prediction accuracy of the value quantity of the commodity object in the target use period in the target purchase period is improved.

In one embodiment of the present disclosure, the method may further include the step of obtaining a time series prediction model, including steps S4100 to S4300 as follows:

step S4100 acquires actual value amounts of the commodity objects in the plurality of training use periods in the corresponding training purchase periods.

The third period time length is the same as the first period time length, and the third period time length is the period time length of the difference between the training use period and the corresponding training purchase period.

In the present embodiment, the cycle duration of the training use cycle, the history use cycle, the target use cycle, the training purchase cycle, the history purchase cycle, and the target purchase cycle is the same.

Step S4200, constructing a first training sample according to the training use period and the actual value amount of the commodity object in the training use period in the corresponding training purchase period.

Wherein the first training sample comprises a matching training time series and an actual value quantity.

In this embodiment, the time length of the training time sequence may be set in advance according to an application scenario or a specific requirement, and the time length may be used to represent the number of corresponding time periods included in the time sequence.

The time length of the training time sequence is the same as the time length of the aforementioned prediction time sequence. Then, the number of training usage periods and corresponding actual value amounts required for constructing one sample may be the same as the number of historical usage periods and corresponding actual value amounts required for predicting the value amount of the commodity object of the target usage period in the target purchase period.

Specifically, the actual value amount included in the first training sample is the actual value amount of the commodity object at the reference training purchase time corresponding to the reference training use time. Wherein, the period duration of the difference between the reference training use time and the training time sequence included in the first training sample is the same as the period duration of the difference between the historical use time and the target use time; the period duration of the difference between the reference training use time and the reference training purchase time is the same as the first period duration and the second period duration in the foregoing embodiment; and the reference training use time is after a training use time included in the training time sequence.

For example, in the embodiment shown in table 4 below in which the commodity object is a ticket for a designated flight, if one training time series includes 7 training use periods and corresponding actual value amounts, and the reference training use time differs from the training time series included in the first training sample by 1 period duration, the training time series may be generated according to the departure date 2019-04-01-2019-04-07 and the price of the corresponding ticket 2 days before the departure (629, 649, 970, 1200, 1475, 540, 660), which may be shown in table 5 below. Correspondingly, the actual value amount matched to the training time series shown in Table 5 may be the price 600 of the ticket for the designated flight with the takeoff date of 2019-04-08 2 days before the takeoff.

TABLE 4

Takeoff date (year-month-day)	Price 2 days before takeoff
		2019-04-01	629
2019-04-02	649
		2019-04-03	970
2019-04-04	1200
		2019-04-05	1475
2019-04-06	540
		2019-04-07	660
2019-04-08	540

TABLE 5

2019-04-01	629
		2019-04-02	649
2019-04-03	970
		2019-04-04	1200
2019-04-05	1475
		2019-04-06	540
2019-04-07	660

And step S4300, based on a preset Prophet algorithm, performing machine learning training according to the first training sample to obtain a time sequence prediction model.

The Prophet algorithm is a time series prediction algorithm and can be used for predicting the future trend of the time series.

In the embodiment, a first training sample is constructed according to the training use period and the actual value quantity of the commodity object in the training use period in the corresponding training purchase period; and based on a preset Prophet algorithm, machine learning training is carried out according to the first training sample, and the value quantity of the commodity object at the target purchasing time in the target using time can be more accurately predicted by the time series prediction model.

In an embodiment of the present disclosure, based on a preset Prophet algorithm, performing machine learning training according to a first training sample, and obtaining a time series prediction model includes steps S4310 to S4330 as follows:

step S4310, determining a predictive value expression of the first training sample by taking the undetermined parameter of the Prophet algorithm as a variable according to the training time sequence of the first training sample.

In one embodiment of the present disclosure, the predicted worth amount expression for the first training sample may be expressed as:

y(t)＝g(t)+s(t)+h(t)+∈

where t is the time information in the training time sequence. g (t) is a trend function representing non-periodic variations over a time series. s (t) the user characterizes periodic variations in the time series. h (t) is a holiday function for fitting holidays and special dates. E is used to fit outliers that cannot be described by the model.

Step S4320, determining an average absolute percentage error between the predicted cost value expression of the first training sample and the actual cost value corresponding to the first training sample as a target function.

For example, the predicted merit expression corresponding to the ith first training sample is y_i(t), the actual value quantity corresponding to the ith first training sample is expressed as y_i ^*In the case where the number of the first training samples is N, the loss function can be expressed as:

in this embodiment, when the value fluctuation range of the commodity object is large, the time series prediction model is optimized by taking the average absolute percentage error as the loss function of the time series prediction model, so that the prediction effect of the time series prediction model can be improved.

And S4330, determining undetermined parameters according to the loss function, and finishing the training of the time series prediction model.

In an embodiment of the present disclosure, a time series prediction model may be obtained by determining the value of a parameter to be determined in a case where the loss function is minimum.

By the time series prediction model obtained by machine learning training according to the first training sample, the value amount of the commodity object at the target purchasing time at the target using time can be predicted more accurately.

And step S3300, a target vector value of the feature vector of the commodity object in the target use period in the target purchase period and a mapping function between the feature vector and the rising and falling scores are obtained.

Wherein the feature vector comprises a plurality of features influencing the rising and falling trend of the commodity object in the corresponding purchase period. The rise and fall score is a score indicating a tendency of the commodity object to rise and fall in the corresponding purchase period.

In one embodiment of the present disclosure, the plurality of features included in the feature vector may be preset according to the type and specific requirements of the commodity object.

For example, where the merchandise object is a ticket for a designated flight, the plurality of features may include: a feature indicating the airline brand of the designated flight, a feature indicating the slot of the designated flight, a feature indicating the remaining number of slots of the designated flight, a feature indicating the passenger capacity of the designated flight, a feature indicating the historical booking status of the designated flight, a feature indicating the departure time of the designated flight, a feature indicating the historical price fluctuation number of the designated flight, a feature indicating the departure date and month of the designated flight, and the like.

For another example, where the merchandise object is a lodging voucher for a designated hotel, the plurality of features can include: a feature indicating a star rating of the designated hotel, a feature indicating a brand of the designated hotel, a feature indicating the number of rooms of the designated hotel, a feature indicating the number of vacant rooms of the designated hotel, and the like.

In an embodiment of the present disclosure, the step of obtaining a mapping function between the feature vector and the rise and fall fraction may include steps S3310 to S3350 as follows:

step S3310, the actual value amount of the commodity object in the training use period in the corresponding training purchase period and the actual value amount in the corresponding next purchase period are obtained.

The third period time length is the same as the first period time length, and the third period time length is the period time length of the difference between the training use period and the corresponding training purchase period. The cycle duration of the training use cycle, the historical use cycle, the target use cycle, the training purchase cycle, the historical purchase cycle and the target purchase cycle are all the same.

The next purchase cycle is the next cycle of the training purchase cycle. Then, in the case where the third period duration is M1, the period duration M2 that is the difference between the training use period and the next purchase period may be: m2 ═ M1-1.

For example, in the embodiment shown in table 6 below in which the merchandise object is a ticket for a specified flight, the departure date is the training use period of the ticket for the specified flight, and 2 days before the departure is the training purchase period of the ticket for the specified flight, then 1 day before the departure may be the next purchase period of the ticket for the specified flight.

TABLE 6

Takeoff date (year-month-day)	Price 1 day before takeoff	2 days before take-offPrice of
			2019-04-01	540	629
2019-04-02	970	649
			2019-04-03	970	970
2019-04-04	1120	1200
			2019-04-05	1480	1475
2019-04-06	450	540
			2019-04-07	930	660
2019-04-08	600	540

Step S3320, comparing the actual value amount of the next purchase cycle with the actual value amount of the training purchase cycle, and determining the actual rising and falling scores of the commodity objects corresponding to the training use cycle according to the comparison result.

Specifically, when the actual value amount of the next purchase period is greater than the actual value amount of the training purchase period, it may be determined that the change trend of the value amount of the commodity object corresponding to the training use period is an increase, and the actual increase/decrease score of the commodity object corresponding to the training use period may be set as the first preset value. When the actual value amount of the next purchase cycle is smaller than the actual value amount of the training purchase cycle, determining that the change trend of the value amount of the commodity object corresponding to the training use cycle is a drop, and may be to set the actual rising and falling score of the commodity object corresponding to the training use cycle as the second preset value. In a case where the actual value amount of the next purchase period is equal to the actual value amount of the training purchase period, it may be determined that the variation trend of the value amount of the commodity object corresponding to the training use period is unchanged, and the actual rising and falling score of the commodity object corresponding to the training use period may be set to a third preset value.

The first preset value, the second preset value and the third preset value are different, and each preset value can be set according to an application scene or specific requirements. For example, the first preset value may be set to 1, the second preset value may be set to 0, and the third preset value may be 0.5.

In the embodiment shown in Table 6 above, for tickets for the designated flight with a takeoff date of 2019-04-01, the price 540 from 1 day before takeoff is less than the price 629 from 2 days before takeoff, so the actual fluctuation score of the tickets for the designated flight of 2019-04-01 can be set to 0. For the ticket for the designated flight with the departure date of 2019-04-02, the price 970 from 1 day before the departure is less than the price 649 from 2 days before the departure, so the actual fluctuation score of the ticket for the designated flight of 2019-04-02 can be set to 1. For a ticket for a designated flight with a departure date of 2019-04-03, the price 970 from 1 day before departure is equal to the price 970 from 2 days before departure, so the actual rise and fall score for the ticket for the designated flight of 2019-04-01 can be set to 0.5.

Step S3330, obtaining training vector values of the feature vectors of the commodity objects in the training use period in the training purchase period.

Step S3340, generating a second training sample according to the training vector value and the actual rise-fall score.

In this embodiment, for any one of the second training samples, the training vector value included is the same as the training use period corresponding to the actual rise-fall score.

Step S3350, training to obtain a mapping function according to the training vector value of the feature vector of the second training sample and the actual rise-fall fraction corresponding to the second training sample.

In this embodiment, the training mode of the mapping function may refer to the aforementioned training mode of the time series prediction model, and is not described herein again.

Through the mapping function obtained by the embodiment, the predicted rising and falling scores of the commodity objects in the target use period in the target purchase period can be more accurate, so that the target predicted value obtained according to the predicted rising and falling scores in the follow-up process is more accurate.

And step S3400, determining the predicted rising and falling scores of the commodity objects in the target use period in the target purchase period according to the target vector value and the mapping function.

The predicted rising and falling score is a score representing the rising and falling trend of the commodity object in the target use period in the target purchase period.

In this embodiment, the target vector value may be input into the mapping function, and the predicted rising and falling score of the commodity object in the target use period within the target purchase period may be obtained.

And step S3500, obtaining the target predicted value quantity of the commodity object in the target purchase period according to the initial predicted value quantity and the predicted rising and falling scores.

In one embodiment of the present disclosure, obtaining the target predicted value amount of the commodity object in the target purchase period according to the initial predicted value amount and the predicted fall-and-rise score may include steps S3510 to S3520 as follows:

in step S3510, the value weight corresponding to the initial predicted value and the score weight corresponding to the predicted rising and falling score are used as the basis.

The value weight and the score weight in this embodiment may be set in advance according to a specific experiment or experience of an engineer.

And S3520, carrying out weighted summation on the initial predicted value quantity and the predicted rising and falling scores according to the value quantity weight and the score weight to obtain a target predicted value quantity of the commodity object in the target use period in the target purchase period.

For example, the value weight may be represented as α, the score weight may be represented as β, and the initial value amount of the merchandise object in the target usage period in the target purchase period may be represented as p_pThe predicted rise and fall score of the commodity object in the target use period in the target purchase period can be represented as p_dThen, the target predicted value amount p of the commodity object of the target usage period in the target purchase period can be expressed as:

p＝α×p_p+β×p_d

according to the embodiment, the initial predicted value quantity and the predicted rising and falling score are subjected to weighted summation according to the preset value quantity weight and the preset score weight, so that the target predicted value quantity of the commodity object in the target use period in the target purchase period is obtained, and the finally obtained target predicted value quantity can be more accurate.

In one embodiment of the present disclosure, the method may further include:

and displaying the target predicted value amount of the commodity object of the target use period in the target purchase period so that the user can select the purchase period of the commodity object of the target use period according to the target predicted value amount.

In the embodiment of the present disclosure, the target purchase cycle may be plural, and the prediction methods of the target predicted value amount of the commodity object in each target purchase cycle for the target usage cycle may be determined through steps S3100 to S3500 of the present disclosure. The time series predictive model used to determine the initial predicted price may be different for different target purchase cycles, as may the mapping function used to determine the predicted rise and fall scores.

In this embodiment, by displaying the target predicted value amount of the commodity object in the target use period in the target purchase period to the user, accurate guidance can be provided for the behavior of the user in purchasing the commodity object. Like this, the user can select the purchase cycle and the life cycle of commodity object according to self demand for the user can also enjoy comparatively preferential price when satisfying self trip demand, promotes user's purchase experience.

< apparatus embodiment >

Corresponding to the method, the disclosure also provides a value quantity prediction device 4000 for the commodity object. As shown in fig. 4, the apparatus 4000 for predicting the value amount of a commodity object may include a first obtaining module 4100, a first predicting module 4200, a second obtaining module 4300, a second predicting module 4400, and a third predicting module 4500.

The first obtaining module 4100 is configured to obtain actual value amounts of the commodity objects in the historical usage periods in the corresponding historical purchase periods.

The first prediction module 4200 is configured to determine an initial predicted value amount of the commodity object in the target purchase cycle in the target usage cycle according to the historical usage cycle and the corresponding actual value amount; the first period time length is the same as the second period time length, the first period time length is the period time length of the difference between the historical using period and the corresponding historical purchasing period, and the second period time length is the period time length of the difference between the target using period and the target ticket purchasing period.

The second obtaining module 4300 is configured to obtain a target vector value of the feature vector of the commodity object in the target use period in the target purchase period, and a mapping function between the feature vector and the fluctuation score; the characteristic vector comprises a plurality of characteristics which influence the rising and falling trend of the commodity object in the corresponding purchase period; the rise and fall score is a score representing a tendency of the commodity object to rise and fall in the corresponding purchase period.

The second prediction module 4400 is configured to determine the predicted rising and falling scores of the commodity objects in the target purchase period according to the target vector value and the mapping function.

The third prediction module 4500 is configured to obtain a target predicted value amount of the commodity object in the target purchase period according to the initial predicted value amount and the predicted rise-and-fall score.

In one embodiment of the present disclosure, the first prediction module 4200 may be further configured to:

and determining the initial prediction value quantity of the commodity object in the target purchase period in the target use period according to the prediction time sequence based on a preset time sequence prediction model.

In one embodiment of the present disclosure, the predicting apparatus 4000 may further include:

the method comprises the steps that under the condition that two adjacent historical use periods are discontinuous, the missing historical use period between the two historical use periods is determined according to the period duration of the historical use period and the period duration of the difference between the two historical use periods;

and the module fills up the missing historical use period between the two historical use periods and marks the actual value quantity of the commodity object with the filled historical use period in the corresponding historical purchase period as a set value.

means for determining a valid range of actual value amounts of the merchandise objects of the historical usage periods in the corresponding historical purchase periods;

a module for culling actual value amounts outside the valid range and corresponding historical usage cycles.

In one embodiment of the disclosure, determining the initial predicted value amount of the commodity object in the target purchase period in the target use period according to the predicted time series based on the preset time series prediction model comprises:

inputting the prediction time sequence into a time sequence prediction model to obtain a prediction result of the value quantity of the commodity object in the target purchase period in the target use period;

and selecting a reference value with the minimum distance from the prediction result as an initial prediction value.

In one embodiment of the disclosure, determining at least one reference value from the actual value quantity comprises:

and selecting the clustering center corresponding to the value quantity classification from the actual value quantities contained in the value quantity classification as a reference value.

the module is used for acquiring the actual value quantity of the commodity objects in the plurality of training use periods in the corresponding training purchase periods; the third period time length is the same as the first period time length, and the third period time length is the period time length of the difference between the training use period and the corresponding training purchase period;

the module is used for constructing a first training sample according to the training use period and the actual value quantity of the commodity object in the training use period in the corresponding training purchase period; the first training sample comprises a training time sequence and an actual value quantity which are matched;

and the module is used for performing machine learning training according to the first training sample based on a preset Prophet algorithm to obtain a time series prediction model.

In an embodiment of the disclosure, based on a preset Prophet algorithm, performing machine learning training according to a first training sample, and obtaining a time series prediction model includes:

determining a predictive value expression of the first training sample by taking undetermined parameters of a Prophet algorithm as variables according to the training time sequence of the first training sample;

determining the average absolute percentage error between the predicted value expression of the first training sample and the actual value corresponding to the first training sample as a target function;

and determining undetermined parameters according to the loss function, and finishing the training of the time series prediction model.

In one embodiment of the present disclosure, the third prediction module 4500 may also be used to:

and carrying out weighted summation on the initial predicted value quantity and the predicted rising and falling scores according to the value quantity weight and the score weight to obtain a target predicted value quantity of the commodity object in the target use period in the target purchase period.

In one embodiment of the present disclosure, the prediction apparatus 4000 may further include a module for obtaining a mapping function between the feature vector and the rise-and-fall score, configured to:

acquiring the actual value quantity of the commodity object in the training use period in the corresponding training purchase period and the actual value quantity in the corresponding next purchase period; wherein the next purchase cycle is the next cycle of the training purchase cycle;

comparing the actual value quantity of the next purchase period with the actual value quantity of the training purchase period, and determining the actual rising and falling scores of the commodity objects corresponding to the training use period according to the comparison result;

acquiring a training vector value of a feature vector of a commodity object in a training use period in a training purchase period;

and training to obtain a mapping function according to the training vector value of the feature vector of the second training sample and the actual rise-fall score corresponding to the second training sample.

and the module is used for displaying the target predicted value quantity of the commodity object of the target use period in the target purchase period so that the user can select the purchase period of the commodity object of the target use period according to the target predicted value quantity.

It will be appreciated by those skilled in the art that the prediction apparatus 4000 of the value amount of the commodity object can be implemented in various ways. For example, the prediction apparatus 4000 of the value amount of the commodity object may be realized by an instruction configuration processor. For example, instructions may be stored in ROM and read from ROM into a programmable device when the device is started up to implement the prediction apparatus 4000 of the value amount of the commodity object. For example, the prediction apparatus 4000 of the value amount of the commodity object may be solidified into a dedicated device (e.g., ASIC). The prediction apparatus 4000 of the value amount of the commodity object may be divided into units independent of each other, or may be implemented by combining them together. The value amount prediction device 4000 for the commodity object may be realized by one of the various implementations described above, or may be realized by a combination of two or more of the various implementations described above.

In this embodiment, the device 4000 for predicting the value amount of the commodity object may have various implementations, for example, the device 4000 for predicting the value amount of the commodity object may be any functional module running in a software product or an application providing a prediction function, or a peripheral insert, a plug-in, a patch, or the like of the software product or the application, or the software product or the application itself.

< electronic device embodiment >

The present disclosure also provides an electronic device 5000. The electronic device 5000 may be a server 1100 as shown in fig. 1. The electronic device 5000 may also be a terminal device 1200 as shown in fig. 2.

In one example, the electronic device 5000 may include the prediction apparatus 4000 for the value amount of the commodity object provided in the foregoing embodiment, which is used to execute the prediction method for the value amount of the commodity object according to any embodiment of the present disclosure.

In another example, as shown in fig. 5, the electronic device 5000 may further include a processor 5100 and a memory 5200, the memory 5200 being for storing computer programs; the computer program is for controlling the processor 5100 to execute the value amount prediction method of the commodity object of any of the embodiments of the present disclosure.

The present disclosure may be systems, methods, and/or computer program products. The computer program product may include a computer-readable storage medium having computer-readable program instructions embodied thereon for causing a processor to implement various aspects of the present disclosure.

The computer readable storage medium may be a tangible device that can hold and store the instructions for use by the instruction execution device. The computer readable storage medium may be, for example, but not limited to, an electronic memory device, a magnetic memory device, an optical memory device, an electromagnetic memory device, a semiconductor memory device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a Static Random Access Memory (SRAM), a portable compact disc read-only memory (CD-ROM), a Digital Versatile Disc (DVD), a memory stick, a floppy disk, a mechanical coding device, such as punch cards or in-groove projection structures having instructions stored thereon, and any suitable combination of the foregoing. Computer-readable storage media as used herein is not to be construed as transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission medium (e.g., optical pulses through a fiber optic cable), or electrical signals transmitted through electrical wires.

The computer-readable program instructions described herein may be downloaded from a computer-readable storage medium to a respective computing/processing device, or to an external computer or external storage device via a network, such as the internet, a local area network, a wide area network, and/or a wireless network. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers and/or edge servers. The network adapter card or network interface in each computing/processing device receives computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in the respective computing/processing device.

The computer program instructions for carrying out operations of the present disclosure may be assembler instructions, Instruction Set Architecture (ISA) instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or secondary code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider). In some embodiments, the electronic circuitry that can execute the computer-readable program instructions implements aspects of the present disclosure by utilizing the state information of the computer-readable program instructions to personalize the electronic circuitry, such as a programmable logic circuit, a Field Programmable Gate Array (FPGA), or a Programmable Logic Array (PLA).

Various aspects of the present disclosure are described herein 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 block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer-readable program instructions.

These computer-readable program instructions may be provided to a processor of a general purpose computer, special purpose computer, 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/acts specified in the flowchart and/or block diagram block or blocks. These computer-readable program instructions may also be stored in a computer-readable storage medium that can direct a computer, programmable data processing apparatus, and/or other devices to function in a particular manner, such that the computer-readable medium storing the instructions comprises an article of manufacture including instructions which implement the function/act specified in the flowchart and/or block diagram block or blocks.

The computer readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other devices to produce a computer implemented process such that the instructions which execute on the computer, other programmable apparatus or other devices implement the functions/acts specified in the flowchart and/or block diagram block or blocks.

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions. It is well known to those skilled in the art that implementation by hardware, by software, and by a combination of software and hardware are equivalent.

Having described embodiments of the present disclosure, the foregoing description is intended to be exemplary, not exhaustive, and not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen in order to best explain the principles of the embodiments, the practical application, or improvements made to the technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein. The scope of the present disclosure is defined by the appended claims.

Claims

1. A method of predicting a value amount of a commodity object, comprising:

2. The method of claim 1, wherein said determining an initial predicted value amount for said commodity object for a target usage period for a target purchase period based on said historical usage periods and corresponding actual value amounts for said historical purchase periods comprises:

3. The method of claim 2, wherein the method further comprises:

4. The method of claim 2, wherein the method further comprises:

5. The method of claim 2, wherein said determining said initial predicted value amount of said commodity object for said target use period for said target purchase period according to said predicted time series based on a preset time series prediction model comprises:

6. The method of claim 5, wherein said determining at least one reference value from said actual value quantity comprises:

7. The method of claim 2, wherein the method further comprises:

8. The method according to claim 7, wherein the performing machine learning training according to the first training sample based on a preset Prophet algorithm to obtain the time series prediction model comprises:

9. The method of claim 1, deriving a target predicted value amount for the commodity object for the target use period within the target purchase period based on the initial predicted value amount and the predicted ramp score comprises:

10. The method of claim 1, wherein the step of obtaining a mapping function between the feature vector and the rise and fall fraction comprises:

11. An apparatus for predicting a value amount of a commodity object, comprising:

12. An electronic device, comprising:

the apparatus of claim 11, or,

a processor and a memory for storing an executable computer program for controlling the processor to perform the method according to any one of claims 1 to 10.