Tobacco demand-based confidence index prediction model training method, apparatus and device, and medium

Abstract

The invention discloses a tobacco demand-based confidence index prediction model training method, apparatus and device, and a medium. The method comprises the following steps: acquiring tobacco index related data in a historical time period; determining a target confidence index value of the corresponding tobacco index related data according to tobacco internal data in the tobacco index related data in the historical time period; according to the tobacco index related data in the historical time period and the corresponding target confidence index value, performing feature screening on the tobacco index related data in the historical time period to obtain a tobacco index training sample; inputting the tobacco index training sample and the target confidence index value corresponding to the tobacco index training sample into a pre-constructed network model to obtain a predicted confidence index value output by the model; performing model training on the network model according to the target confidence index value and the predicted confidence index value to obtain a confidence index prediction model; the confidence index prediction model is used for predicting the tobacco demand confidence index in the future time period.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method, device, equipment and medium for training a confidence index prediction model based on tobacco demand. Background Art

[0002] In order to effectively help enterprises or the public understand the market trend of the cigarette market, the confidence index of the cigarette market can be predicted to identify potential opportunities and risks in the cigarette market, so as to further formulate relevant policies for the cigarette market. Therefore, it is necessary to provide a method for predicting the confidence index of the cigarette market to accurately predict the confidence index of the cigarette market, so as to help relevant practitioners gain a deeper understanding of the future development trend of the cigarette market. Summary of the invention

[0003] The present invention provides a tobacco demand-based confidence index prediction model training method, device, equipment and medium to achieve accurate prediction of the tobacco market confidence index.

[0004] According to one aspect of the present invention, a method for training a confidence index prediction model based on tobacco demand is provided, the method comprising:

[0005] Obtain tobacco index related data in historical time periods;

[0006] Determining a target confidence index value of the corresponding tobacco index-related data according to the tobacco internal data in the tobacco index-related data in the historical time period;

[0007] According to the tobacco index related data in the historical time period and the corresponding target confidence index value, feature screening is performed on the tobacco index related data in the historical time period to obtain tobacco index training samples;

[0008] Inputting the tobacco index training sample and its corresponding target confidence index value into a pre-built network model to obtain a prediction confidence index value output by the model;

[0009] The network model is trained according to the target confidence index value and the predicted confidence index value to obtain a confidence index prediction model; the confidence index prediction model is used to predict the tobacco demand confidence index in a future time period.

[0010] According to another aspect of the present invention, a method for predicting a confidence index based on tobacco demand is provided, comprising:

[0011] Determine the forecast demand for the tobacco market confidence index;

[0012] According to the forecast demand of the tobacco market confidence index, obtain tobacco index related data in the historical time period;

[0013] Inputting the tobacco index related data into a pre-trained confidence index prediction model to obtain a prediction confidence index for a future time period output by the model;

[0014] The confidence index prediction model is generated by using the confidence index prediction model training method based on tobacco demand as described in the above embodiment.

[0015] According to another aspect of the present invention, a tobacco demand-based confidence index prediction model training device is provided, the device comprising:

[0016] A data acquisition module is used to obtain tobacco index related data in historical time periods;

[0017] An index determination module, used to determine a target confidence index value of corresponding tobacco index-related data according to tobacco internal data in the tobacco index-related data in the historical time period;

[0018] A sample generation module, for performing feature screening on the tobacco index related data in the historical time period according to the tobacco index related data in the historical time period and the corresponding target confidence index value, to obtain tobacco index training samples;

[0019] An index prediction module, used to input the tobacco index training sample and its corresponding target confidence index value into a pre-built network model to obtain a prediction confidence index value output by the model;

[0020] The model training module is used to perform model training on the network model according to the target confidence index value and the predicted confidence index value to obtain a confidence index prediction model; the confidence index prediction model is used to predict the tobacco demand confidence index in a future time period.

[0021] According to another aspect of the present invention, a device for predicting a confidence index based on tobacco demand is provided, the device comprising:

[0022] Demand determination module, used to determine the forecast demand of tobacco market confidence index;

[0023] A data acquisition module, used to predict demand according to the tobacco market confidence index and acquire tobacco index related data in a historical time period;

[0024] An index prediction module, used to input the tobacco index related data into a pre-trained confidence index prediction model to obtain a prediction confidence index output by the model for a future time period;

[0025] The confidence index prediction model is generated by using the confidence index prediction model training method based on tobacco demand as described in the above embodiment.

[0026] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0027] at least one processor; and

[0028] a memory communicatively connected to the at least one processor; wherein,

[0029] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the confidence index prediction model training based on tobacco demand and / or the confidence index prediction method based on tobacco demand described in any embodiment of the present invention.

[0030] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the confidence index prediction model training based on tobacco demand and / or the confidence index prediction method based on tobacco demand described in any embodiment of the present invention when executed.

[0031] The technical solution of the embodiment of the present invention determines the target confidence index value of the corresponding tobacco index related data according to the tobacco internal data in the tobacco index related data under the historical time period; performs feature screening on the tobacco index related data under the historical time period according to the tobacco index related data under the historical time period and the corresponding target confidence index value to obtain tobacco index training samples; inputs the tobacco index training samples and their corresponding target confidence index values ​​into a pre-constructed network model to obtain the predicted confidence index value output by the model; performs model training on the network model according to the target confidence index value and the predicted confidence index value to obtain a confidence index prediction model to predict the tobacco demand confidence index under the future time period. The above technical solution combines the tobacco index related data of each dimension, constructs the sample data, and performs feature screening on the feature similarity between the data of each dimension, thereby improving the model training accuracy of the tobacco index prediction model, thereby improving the accurate prediction of the tobacco market confidence index.

[0032] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0034] Figure 1 is a flowchart of a confidence index prediction model training method based on tobacco demand provided in accordance with the first embodiment of the present invention;

[0035] Figure 2 is a flow chart of a confidence index prediction method based on tobacco demand provided in accordance with Embodiment 2 of the present invention;

[0036] Figure 3 It is a structural schematic diagram of a confidence index prediction model training device based on tobacco demand provided according to Embodiment 3 of the present invention;

[0037] Figure 4 is a schematic structural diagram of a confidence index prediction device based on tobacco demand provided according to Embodiment 4 of the present invention;

[0038] Figure 5 It is a structural schematic diagram of an electronic device for implementing the confidence index prediction model training based on tobacco demand and / or the confidence index prediction method based on tobacco demand according to an embodiment of the present invention. DETAILED DESCRIPTION

[0039] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0040] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0041] Embodiment 1

[0042] Figure 1 A flowchart of a confidence index prediction model training method based on tobacco demand is provided in Example 1 of the present invention. This embodiment is applicable to the situation of predicting the tobacco market confidence index. The method can be executed by a confidence index prediction model training device based on tobacco demand. The confidence index prediction model training device based on tobacco demand can be implemented in the form of hardware and / or software. The confidence index prediction model training device based on tobacco demand can be configured in an electronic device.

[0043] The acquisition, storage, use, and processing of data in the technical solution of the present invention are in compliance with the relevant provisions of national laws and regulations. It should be noted that in this embodiment, some existing solutions in the industry such as certain software, components, and models may be mentioned, which should be considered as exemplary, and their purpose is only to illustrate the feasibility of the implementation of the technical solution of this application, but it does not mean that the applicant has or will necessarily use the relevant content of the solution.

[0044] like Figure 1 A flowchart of a confidence index prediction model training method based on tobacco demand is shown, the method comprising:

[0045] S110. Obtain tobacco index related data in a historical time period.

[0046] S120. Determine a target confidence index value of corresponding tobacco index-related data according to tobacco internal data in tobacco index-related data in a historical time period.

[0047] S130. Based on the tobacco index related data in the historical time period and the corresponding target confidence index value, feature screening is performed on the tobacco index related data in the historical time period to obtain tobacco index training samples.

[0048] S140, inputting tobacco index training samples and their corresponding target confidence index values ​​into a pre-built network model to obtain a predicted confidence index value output by the model.

[0049] S150. According to the target confidence index value and the predicted confidence index value, the network model is trained to obtain a confidence index prediction model; the confidence index prediction model is used to predict the tobacco demand confidence index in the future time period.

[0050] Among them, tobacco index related data may include tobacco internal data and tobacco external data; tobacco external data may include network signaling data, resource macro data and Internet search index data; tobacco internal data may include tobacco resource demand data and tobacco trading equipment data. Among them, network signaling data, resource macro data, Internet search index data, tobacco resource demand data and tobacco trading equipment data all include relevant index data of at least one dimension.

[0051] Specifically, mobile phone signaling data may include relevant index data such as total population data, visiting population data, working population data, youth population data and middle-aged male population data related to the cigarette market; resource macro data may include relevant index data such as industrial growth value, total retail sales of consumer goods, product price fluctuations and market share; Internet search index data may include personnel income data, employment-related data and tobacco index-related data; tobacco resource demand data may include relevant index data such as actual demand of retailers, actual purchase volume of retailers, average price of cigarette strips, order frequency of retailers, order satisfaction rate and actual purchase amount of retailers; tobacco trading equipment data may include relevant index data such as tobacco demand, tobacco value attribute value, order volume of typical specifications of cigarettes and order price of typical specifications.

[0052] The historical time period may be the cigarette index-related data for each month in the past 24 months. The specific time may be pre-set by relevant technical personnel according to actual needs; the time period may be based on a month.

[0053] Among them, the confidence index value is used to characterize the confidence index in the future cycle of the tobacco market. On the one hand, it reflects the consumer confidence status of the cigarette market, and on the other hand, it can reflect the future consumption trend of the cigarette market. The higher the confidence index value, the higher the confidence in the future of the tobacco market; the lower the confidence index value, the lower the confidence in the future of the tobacco market. The construction method of the confidence index function of tobacco index related data can be based on the existing Xuwen Pineapple Index, and can also be constructed based on the consumer confidence index. This embodiment combines the characteristics of the cigarette market and is constructed based on tobacco demand and tobacco value attribute values.

[0054] In an optional embodiment, the tobacco transaction device data includes tobacco demand and tobacco value attribute values; the tobacco index-related data in the historical time period is tobacco index-related data at at least one historical time node; accordingly, according to the tobacco internal data in the tobacco index-related data in the historical time period, determining the target confidence index value of the corresponding tobacco index-related data includes:

[0055] Step a11: Determine the mean value and standard deviation of tobacco demand based on the tobacco demand.

[0056] Specifically, the tobacco demand in the historical time period is averaged to obtain the mean demand; the standard deviation of the demand is determined based on the tobacco demand in the historical time period and the mean demand.

[0057] Step a12: Determine the value mean and value standard deviation based on the tobacco value attribute value.

[0058] Among them, the average value of tobacco value attributes in the historical time period is calculated to obtain the mean value; based on the value attribute values ​​and the mean value of each tobacco in the historical time period, the value standard deviation is determined.

[0059] Step a2: For any historical time node, according to the tobacco demand and tobacco value attribute values ​​at the historical time node, and the tobacco demand and tobacco price attribute values ​​at the predicted time node corresponding to the historical time node, based on the demand mean, demand standard deviation, value mean and value standard deviation, determine the target confidence index value of the tobacco index-related data at the historical node.

[0060] Specifically, the confidence index prediction function constructed in this embodiment is expressed as follows:

[0061]

[0062] Among them, P t represents the tobacco demand at the historical time node t; P t+1 represents the tobacco demand at the predicted time node corresponding to the historical time node t; represents the mean value of demand; σ P Indicates the standard deviation of demand; Sale t Represents the tobacco price attribute value at the historical time node t; Sale t+1 Represents the tobacco price attribute value at the predicted time node corresponding to the historical time node t; represents the mean value; σ sale represents the standard deviation of value; Δ represents the random impact factor that has an additional impact on the confidence index value; the values ​​of α, β, γ and δ can be pre-set by relevant technical personnel based on actual experience or experiments. Preferably, α, β, γ and v are set to 0.2, 0.3, 0.2 and 0.3 respectively; when there is a random impact index, the value of Δ can be pre-calculated by relevant technical personnel; when there is no random impact index, the value of Δ is zero. For example, if time t is April 15, 2024, then P t+1 It can be May 15, 2024.

[0063] For any tobacco index-related data, its corresponding target confidence index value can be determined. During the model training process, the tobacco index-related data is used as a sample training set, and the target confidence index value can be used as the label value of the corresponding tobacco index-related data.

[0064] It should be noted that in order to further ensure the diversity, compliance, comprehensiveness and standardization of sample generation, the tobacco index-related data can be further preprocessed and screened for data features to obtain tobacco index training samples for model training.

[0065] The tobacco index-related data were processed for noise, outliers and missing values, the data with different dimensions were standardized, the duplicate data were deduplicated, the fields with less missing data were smoothed, and the fields with more missing data were eliminated.

[0066] For different types of tobacco index-related data, such as network signaling data and tobacco resource demand data, clarify the specific meaning of each field or variable in the data set, check the value range and type of the data in each field, and determine whether there are missing values ​​or abnormal values. Analyze the distribution of tobacco index-related data in different regions. There are more or less differences in different regions. Therefore, it is necessary to consider regional comprehensiveness and include tobacco index-related data in multiple regions.

[0067] In order to achieve the comprehensiveness of sample data, it is necessary to construct the sample data set from multiple dimensions. Taking tobacco index-related data as raw data, additional features are inferred or constructed from the original features of the raw data through feature combination or feature segmentation. For example, based on youth population data and male population data, young male population data can be obtained by fusion; for another example, based on household order volume, household order amount, order frequency and purchase volume, additional features such as average household order volume, average household order amount, order frequency and average household planned purchase volume can be obtained by fusion.

[0068] Since the cigarette scene is affected by time, the correlation or association between different samples at different times, as well as the correlation between tobacco index related data and the confidence index value, can be considered. Various types of tobacco index related data can also be screened in the time dimension and correlation dimension to obtain tobacco index training samples.

[0069] In an optional embodiment, based on the tobacco index related data in the historical time period and the corresponding target confidence index value, feature screening is performed on the tobacco index related data in the historical time period to obtain tobacco index training samples, including:

[0070] Step b1, generating relevant index trend curves corresponding to the relevant index data in the historical time period; and generating a confidence index trend curve corresponding to the target confidence index value in the historical time period.

[0071] For any relevant index data under any tobacco index related data, for example, the relevant index data is the actual purchase volume of the retailer, and according to the actual purchase volume of the retailer in the historical time period, a confidence index curve of the actual purchase volume of the retailer is generated. For example, taking 24 months as an example, a corresponding confidence index curve is generated according to the actual purchase volume of the retailer in each month.

[0072] Similarly, based on the target confidence index under the historical time period, a corresponding confidence index trend curve is generated.

[0073] Step b2: determining the characteristic correlation between each correlation index trend curve and the confidence index trend curve.

[0074] Exemplarily, the characteristic correlation between each correlation index trend curve and the confidence index trend curve can be determined based on a preset correlation calculation algorithm, wherein the correlation calculation algorithm can be, for example, Pearson Correlation Coefficient, Spearman's Rank Correlation Coefficient, Curve Fitting or Cross-Correlation Function (CCF).

[0075] Step b3: Based on the correlation of each feature, feature screening is performed on tobacco index related data in the historical time period to obtain tobacco index training samples.

[0076] Specifically, the correlation index data corresponding to the correlation index trend curve whose characteristic correlation is greater than a preset correlation threshold can be used as tobacco index training samples, wherein the correlation threshold can be preset by relevant technical personnel.

[0077] It should be noted that due to the temporal characteristics of tobacco data, there may be a correlation between two tobacco indices in time. For example, the data of Index A from June to October is correlated with the data of Index B from January to May, but the data of Index A from January to May may not be highly correlated with the data of Index B from January to May in the same time period. Therefore, Index A can be time-series shifted to obtain index data with a higher correlation with Index B.

[0078] Optionally, before generating the relevant index trend curves corresponding to the relevant index data in the historical time period, it also includes: performing time series shift on the relevant index data to obtain the translation index data corresponding to the relevant index data; performing data screening on the relevant index data according to the data feature similarity between the relevant index data and the corresponding translation index data to obtain the screened relevant index data; accordingly, generating the relevant index trend curves corresponding to the relevant index data in the historical time period, including: generating the relevant index trend curves corresponding to the screened relevant index data in the historical time period.

[0079] The time series range of the translation can be preset by relevant technical personnel according to actual needs. For any relevant index data, the number of translation index data obtained by time series translation of the relevant index data can be one or more. If there are multiple translation index data, the relevant index data can be obtained by time series translation based on different preset time series ranges.

[0080] Determine the data feature similarity between each related index data and between the translation index data. For example, if the related index data a has translation index data a1, the related index data b has translation index data b1, and the related index data c has translation index data c1, then determine the data feature similarity between the related index data a and the related feature index b, translation index data b1, related index data c, and translation index data c1, respectively, and determine the data feature similarity between the related index data a1 and the related feature index b, translation index data b1, related index data c, and translation index data c1, respectively. The data feature similarity calculation method between other related index data is similar, and this embodiment will not be repeated.

[0081] According to the data feature similarity, each related index data is screened to obtain the screened related index data. For example, if the data feature similarity between the translation index data a and the related index data b is not greater than the preset similarity threshold, but the data feature similarity between the translation index data a1 and the related index data b is greater than the preset similarity threshold, then the related index data a and the related index data b can be retained. If the data feature similarity between the translation index data a and the related index data b is not greater than the preset similarity threshold, and the data feature similarity between the translation index data a1 and the related index data b is also not greater than the preset similarity threshold, then the related index data a and the related index data b can be eliminated. Generate the related index trend curves corresponding to each of the screened related index data under the historical time period.

[0082] The tobacco index training samples and their corresponding target confidence index values ​​are input into the pre-built network model to obtain the predicted confidence index value output by the model.

[0083] Among them, the network model can be pre-selected by relevant technical personnel according to the actual tobacco scene prediction needs. Preferably, the network model can be a random forest model and an ARIMA (Autoregressive Integrated Moving Average). Among them, the ARIMA model is a classic model for time series analysis and prediction, which combines autoregression (AR) and moving average (MA), specifically used to describe trends, seasonality and randomness in time series data.

[0084] The tobacco index training samples and their corresponding target confidence index values ​​are input into a pre-built network model, and the model predicts the confidence index values ​​of the tobacco index training samples to obtain the predicted confidence index values ​​output by the model.

[0085] According to the target confidence index value and predicted confidence index value of the tobacco index training sample, the network model is trained to obtain a confidence index prediction model; wherein the confidence index prediction model is used to predict the tobacco demand confidence index in the future time period. Specifically, according to the target confidence index value and the predicted confidence index value, based on the preset loss function, the target loss value can be determined, and it can be judged whether the target loss value tends to be stable. If so, the model training is terminated to obtain the confidence index prediction model.

[0086] Among them, the confidence index prediction model is used to predict the tobacco demand confidence index in the future time period. For example, it is expected to predict the tobacco confidence index in the future time period of October 2024. Then, the tobacco-related index data from January 2024 to September 2024 in the historical time period can be used as input to the confidence index prediction model, and the model outputs the tobacco confidence index in the future time period of October 2024.

[0087] It should be noted that the benchmark value of the tobacco confidence index can be set in advance. The determination of the benchmark value can adopt the fixed-base index method, including fixed-node fixed-base (focusing on the long term) and month-on-month fixed-base (focusing on the short term). Considering the long-term effect of the index and the actual situation of the cigarette market, this model can adopt the fixed-node fixed-base method, with the average confidence index from June 2023 to December 2023 as 100, and the overall cigarette market confidence index is scaled and translated to determine the specific value of the confidence index.

[0088] The technical solution of the embodiment of the present invention determines the target confidence index value of the corresponding tobacco index related data according to the tobacco internal data in the tobacco index related data under the historical time period; performs feature screening on the tobacco index related data under the historical time period according to the tobacco index related data under the historical time period and the corresponding target confidence index value to obtain tobacco index training samples; inputs the tobacco index training samples and their corresponding target confidence index values ​​into a pre-constructed network model to obtain the predicted confidence index value output by the model; performs model training on the network model according to the target confidence index value and the predicted confidence index value to obtain a confidence index prediction model to predict the tobacco demand confidence index under the future time period. The above technical solution combines the tobacco index related data of each dimension, constructs the sample data, and performs feature screening on the feature similarity between the data of each dimension, thereby improving the model training accuracy of the tobacco index prediction model, thereby improving the accurate prediction of the tobacco market confidence index.

[0089] Embodiment 2

[0090] Figure 2 This is a flowchart of a confidence index prediction method based on tobacco demand provided in the second embodiment of the present invention. This embodiment is applicable to the case of predicting the confidence index of the tobacco market. The method can be executed by a confidence index prediction device based on tobacco demand. The confidence index prediction device based on tobacco demand can be implemented in the form of hardware and / or software. The confidence index prediction device based on tobacco demand can be configured in an electronic device. Figure 2 As shown, the method includes:

[0091] S210. Determine the forecast demand for the tobacco market confidence index.

[0092] S220. Predict demand based on the tobacco market confidence index and obtain tobacco index related data in a historical time period.

[0093] S230, inputting tobacco index related data into a pre-trained confidence index prediction model to obtain a prediction confidence index for a future time period output by the model.

[0094] The confidence index prediction model is generated by using the confidence index prediction model training method based on tobacco demand as described in the above embodiment.

[0095] It should be noted that the prediction of the tobacco market confidence index can be made from at least three dimensions, for example, including the regional dimension, the price category dimension and the industry dimension. Among them, the regional dimension can be used to predict the tobacco market confidence index of a certain region in a certain month (or several months) in the future; the price category dimension can be used to predict the tobacco market confidence index of a certain price category in a certain month in the future; the industry dimension can be used to predict the tobacco market confidence index of a certain tobacco industry in a certain month in the future.

[0096] The demand for tobacco market confidence index prediction may be a demand for tobacco index prediction in a certain dimension, for example, it may be a prediction demand in a regional dimension, it may be a prediction demand in an industrial dimension, or it may be a prediction demand combining a regional dimension and a price category dimension.

[0097] According to different tobacco market confidence index forecasting needs, tobacco index related data in the historical period corresponding to the forecasting needs can be obtained. The tobacco index related data is input into the pre-trained confidence index forecasting model to obtain the forecast confidence index in the future time period output by the model.

[0098] For example, if the tobacco market confidence index prediction demand is to predict the tobacco market confidence index of a certain region, and the prediction target is to predict the tobacco market confidence index of region A in October 2024, then tobacco index-related data under different dimensions related to region A from January 2024 to September 2024 can be obtained. The tobacco index-related data is input into the confidence index prediction model based on the above embodiment, and the model's prediction confidence index for region A in October 2024 can be obtained.

[0099] The technical solution of this embodiment determines the demand for tobacco market confidence index prediction, obtains tobacco index related data in historical time periods based on the tobacco market confidence index prediction demand, inputs the tobacco index related data into a pre-trained confidence index prediction model, obtains the predicted confidence index in the future time period output by the model, and realizes accurate prediction of the tobacco market confidence index.

[0100] Embodiment 3

[0101] Figure 3 A schematic diagram of a tobacco demand-based confidence index prediction model training device provided in the third embodiment of the present invention. A tobacco demand-based confidence index prediction model training device provided in the embodiment of the present invention can be applied to the situation of predicting the tobacco market confidence index. The tobacco demand-based confidence index prediction model training device can be implemented in the form of hardware and / or software, such as Figure 3As shown, the device specifically includes: a data acquisition module 301, an index determination module 302, a sample generation module 303, an index prediction module 304 and a model training module 305.

[0102] in,

[0103] The data acquisition module 301 is used to acquire tobacco index related data in a historical time period;

[0104] An index determination module 302, for determining a target confidence index value of corresponding tobacco index-related data according to tobacco internal data in the tobacco index-related data in the historical time period;

[0105] The sample generation module 303 is used to perform feature screening on the tobacco index related data in the historical time period according to the tobacco index related data in the historical time period and the corresponding target confidence index value to obtain tobacco index training samples;

[0106] The index prediction module 304 is used to input the tobacco index training sample and its corresponding target confidence index value into a pre-built network model to obtain a prediction confidence index value output by the model;

[0107] The model training module 305 is used to perform model training on the network model according to the target confidence index value and the predicted confidence index value to obtain a confidence index prediction model; the confidence index prediction model is used to predict the tobacco demand confidence index in the future time period.

[0108] The technical solution of the embodiment of the present invention determines the target confidence index value of the corresponding tobacco index related data according to the tobacco internal data in the tobacco index related data under the historical time period; performs feature screening on the tobacco index related data under the historical time period according to the tobacco index related data under the historical time period and the corresponding target confidence index value to obtain tobacco index training samples; inputs the tobacco index training samples and their corresponding target confidence index values ​​into a pre-constructed network model to obtain the predicted confidence index value output by the model; performs model training on the network model according to the target confidence index value and the predicted confidence index value to obtain a confidence index prediction model to predict the tobacco demand confidence index under the future time period. The above technical solution combines the tobacco index related data of each dimension, constructs the sample data, and performs feature screening on the feature similarity between the data of each dimension, thereby improving the model training accuracy of the tobacco index prediction model, thereby improving the accurate prediction of the tobacco market confidence index.

[0109] Optionally, the tobacco index-related data includes tobacco internal data and tobacco external data; the tobacco external data includes network signaling data, resource macro data and Internet search index data; the tobacco internal data includes tobacco resource demand data and tobacco trading equipment data; the network signaling data, resource macro data, Internet search index data, tobacco resource demand data and tobacco trading equipment data all include relevant index data of at least one dimension.

[0110] Optionally, the tobacco transaction device data includes tobacco demand and tobacco value attribute value; the tobacco index related data in the historical time period is tobacco index related data in at least one historical time node;

[0111] Accordingly, the index determination module 302 is specifically configured to:

[0112] Determining a mean value and a standard deviation of the demand based on the tobacco demand; and,

[0113] Determining a value mean and a value standard deviation based on the tobacco value attribute value;

[0114] For any historical time node, according to the tobacco demand and tobacco value attribute values ​​at the historical time node, and the tobacco demand and tobacco price attribute values ​​at the predicted time node corresponding to the historical time node, based on the demand mean, demand standard deviation, value mean and value standard deviation, determine the target confidence index value of the tobacco index-related data of the historical node.

[0115] Optionally, the sample generation module 303 includes:

[0116] An index trend curve generating unit, used to generate relevant index trend curves corresponding to the relevant index data in the historical time period; and to generate a confidence index trend curve corresponding to the target confidence index value in the historical time period;

[0117] A correlation determination unit, used to determine the characteristic correlation between each of the correlation index trend curves and the confidence index trend curve;

[0118] The feature screening unit is used to perform feature screening on tobacco index related data in a historical time period according to the feature correlations to obtain tobacco index training samples.

[0119] Optionally, the sample generation module 303 further includes:

[0120] A translation index data generating unit, used for performing time series translation on each of the related index data to obtain translation index data corresponding to each of the related index data before generating the related index trend curves corresponding to each of the related index data in the historical time period;

[0121] A data screening unit, used for screening each of the related index data according to the data feature similarity between each of the related index data and its corresponding translation index data, to obtain screened related index data;

[0122] Accordingly, an exponential trend curve generating unit is generated, which is specifically used for:

[0123] Generate relevant index trend curves corresponding to each of the filtered relevant index data in a historical time period.

[0124] The tobacco demand-based confidence index prediction model training device provided in the embodiment of the present invention can execute the tobacco demand-based confidence index prediction model training method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0125] Embodiment 4

[0126] Figure 4 A schematic diagram of a confidence index prediction device based on tobacco demand provided in the fourth embodiment of the present invention. A confidence index prediction device based on tobacco demand provided in the embodiment of the present invention can be applied to the situation of predicting the confidence index of the tobacco market. The confidence index prediction device based on tobacco demand can be implemented in the form of hardware and / or software, such as Figure 4 As shown, the device specifically includes: a demand determination module 401, a data acquisition module 402 and an index prediction module 403. Among them,

[0127] Demand determination module 401, used to determine tobacco market confidence index forecast demand;

[0128] The data acquisition module 402 is used to predict demand according to the tobacco market confidence index and acquire tobacco index related data in a historical time period;

[0129] The index prediction module 403 is used to input the tobacco index related data into the pre-trained confidence index prediction model to obtain the prediction confidence index output by the model in the future time period;

[0130] The confidence index prediction model is generated by using the confidence index prediction model training method based on tobacco demand as described in the above embodiment.

[0131] The technical solution of this embodiment determines the demand for tobacco market confidence index prediction, obtains tobacco index related data in historical time periods based on the tobacco market confidence index prediction demand, inputs the tobacco index related data into a pre-trained confidence index prediction model, obtains the predicted confidence index in the future time period output by the model, and realizes accurate prediction of the tobacco market confidence index.

[0132] The tobacco demand-based confidence index prediction model training device provided in the embodiment of the present invention can execute the tobacco demand-based confidence index prediction model training method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0133] Embodiment 5

[0134] Figure 5 A schematic diagram of an electronic device 50 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0135] like Figure 5 As shown, the electronic device 50 includes at least one processor 51, and a memory connected to the at least one processor 51 in communication, such as a read-only memory (ROM) 52, a random access memory (RAM) 53, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 51 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 52 or the computer program loaded from the storage unit 58 to the random access memory (RAM) 53. In the RAM 53, various programs and data required for the operation of the electronic device 50 can also be stored. The processor 51, the ROM 52, and the RAM 53 are connected to each other via a bus 54. An input / output (I / O) interface 55 is also connected to the bus 54.

[0136] A number of components in the electronic device 50 are connected to the I / O interface 55, including: an input unit 56, such as a keyboard, a mouse, etc.; an output unit 57, such as various types of displays, speakers, etc.; a storage unit 58, such as a disk, an optical disk, etc.; and a communication unit 59, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 59 allows the electronic device 50 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0137] The processor 51 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 51 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The processor 51 executes the various methods and processes described above, such as confidence index prediction model training based on tobacco demand and / or confidence index prediction method based on tobacco demand.

[0138] In some embodiments, the confidence index prediction model training based on tobacco demand and / or the confidence index prediction method based on tobacco demand can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 58. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 50 via the ROM 52 and / or the communication unit 59. When the computer program is loaded into the RAM 53 and executed by the processor 51, one or more steps of the confidence index prediction model training based on tobacco demand and / or the confidence index prediction method based on tobacco demand described above can be performed. Alternatively, in other embodiments, the processor 51 can be configured to perform the confidence index prediction model training based on tobacco demand and / or the confidence index prediction method based on tobacco demand by any other appropriate means (e.g., by means of firmware).

[0139] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0140] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0141] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0142] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0143] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0144] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0145] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0146] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A confidence index prediction model training method based on tobacco demand, characterized in that: include: Obtain tobacco index related data in historical time periods; Determining a target confidence index value of the corresponding tobacco index-related data according to the tobacco internal data in the tobacco index-related data in the historical time period; According to the tobacco index related data in the historical time period and the corresponding target confidence index value, feature screening is performed on the tobacco index related data in the historical time period to obtain tobacco index training samples; Inputting the tobacco index training sample and its corresponding target confidence index value into a pre-built network model to obtain a prediction confidence index value output by the model; The network model is trained according to the target confidence index value and the predicted confidence index value to obtain a confidence index prediction model; the confidence index prediction model is used to predict the tobacco demand confidence index in a future time period.

2. The method according to claim 1, characterized in that The tobacco index-related data include tobacco internal data and tobacco external data; the tobacco external data include network signaling data, resource macro data and Internet search index data; the tobacco internal data include tobacco resource demand data and tobacco trading equipment data; the network signaling data, resource macro data, Internet search index data, tobacco resource demand data and tobacco trading equipment data all include relevant index data of at least one dimension.

3. The method according to claim 2, characterized in that The tobacco transaction device data includes tobacco demand and tobacco value attribute value; the tobacco index related data under the historical time period is tobacco index related data under at least one historical time node; Accordingly, determining the target confidence index value of the corresponding tobacco index-related data according to the tobacco internal data in the tobacco index-related data in the historical time period includes: Determining a mean value and a standard deviation of the demand based on the tobacco demand; and, Determining a value mean and a value standard deviation based on the tobacco value attribute value; For any historical time node, according to the tobacco demand and tobacco value attribute values ​​at the historical time node, and the tobacco demand and tobacco price attribute values ​​at the predicted time node corresponding to the historical time node, based on the demand mean, demand standard deviation, value mean and value standard deviation, determine the target confidence index value of the tobacco index-related data of the historical node.

4. The method according to claim 2, characterized in that: The method of performing feature screening on the tobacco index related data in the historical time period according to the tobacco index related data in the historical time period and the corresponding target confidence index value to obtain tobacco index training samples includes: Generate relevant index trend curves corresponding to the relevant index data in the historical time period; and generate confidence index trend curves corresponding to the target confidence index values ​​in the historical time period; Determining characteristic correlations between each of the correlation index trend curves and the confidence index trend curve; According to the correlation of each feature, feature screening is performed on tobacco index related data in a historical time period to obtain tobacco index training samples.

5. The method according to claim 4, characterized in that Before generating the relevant index trend curves corresponding to the relevant index data in the historical time period, the method further includes: Performing time series translation on each of the related index data to obtain translation index data corresponding to each of the related index data; According to the data feature similarity between each of the related index data and its corresponding translation index data, data screening is performed on each of the related index data to obtain screened related index data; Accordingly, the relevant index trend curves corresponding to the relevant index data in the historical time period are generated, including: Generate relevant index trend curves corresponding to each of the filtered relevant index data in a historical time period.

6. A confidence index prediction method based on tobacco demand, characterized in that: include: Determine the forecast demand for the tobacco market confidence index; According to the forecast demand of the tobacco market confidence index, obtain tobacco index related data in the historical time period; Inputting the tobacco index related data into a pre-trained confidence index prediction model to obtain a prediction confidence index for a future time period output by the model; Wherein, the confidence index prediction model is generated by using the confidence index prediction model training method based on tobacco demand as described in any one of claims 1 to 5.

7. A confidence index prediction model training device based on tobacco demand, characterized in that: include: A data acquisition module is used to obtain tobacco index related data in historical time periods; An index determination module, used to determine a target confidence index value of corresponding tobacco index-related data according to tobacco internal data in the tobacco index-related data in the historical time period; A sample generation module, for performing feature screening on the tobacco index related data in the historical time period according to the tobacco index related data in the historical time period and the corresponding target confidence index value, to obtain tobacco index training samples; An index prediction module, used to input the tobacco index training sample and its corresponding target confidence index value into a pre-built network model to obtain a prediction confidence index value output by the model; The model training module is used to perform model training on the network model according to the target confidence index value and the predicted confidence index value to obtain a confidence index prediction model; the confidence index prediction model is used to predict the tobacco demand confidence index in a future time period.

8. A confidence index prediction device based on tobacco demand, characterized in that: include: Demand determination module, used to determine the forecast demand of tobacco market confidence index; A data acquisition module, used to predict demand according to the tobacco market confidence index and acquire tobacco index related data in a historical time period; An index prediction module, used to input the tobacco index related data into a pre-trained confidence index prediction model to obtain a prediction confidence index output by the model for a future time period; Wherein, the confidence index prediction model is generated by using the confidence index prediction model training method based on tobacco demand as described in any one of claims 1 to 5.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the method according to any one of claims 1 to 5 and / or claim 6.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method of any one of claims 1-5 and / or claim 6 when executed.

Images