Prediction method based on multi-channel sequence data and related equipment

Abstract

The invention provides a multi-channel sequence data-based prediction method and related equipment. The method comprises the following steps: acquiring multi-channel sequence data related to a prediction task specified by a user; the multi-channel sequence data comprises state information of a prediction object associated with the prediction task at a plurality of past time points; performing multi-channel adaptive decomposition operation on the multi-channel sequence data to obtain main frequency components and dynamic characteristics in the multi-channel sequence data; the main frequency components and the dynamic features are input into a prediction model, output data of the prediction model are obtained to serve as a prediction result, and the prediction result comprises state information of the prediction object at multiple time points in the future. According to the method, higher-quality features are provided for the prediction model through multi-channel adaptive Fourier decomposition, the prediction precision can be improved, dependence on a long-time sequence is effectively reduced through combination of the multi-channel adaptive decomposition and the long-short-term memory network, and the gradient disappearance problem in long-time sequence prediction is relieved.

Description

Technical Field

[0001] The present application belongs to the technical field of signal processing and prediction analysis, and in particular, relates to a method and related equipment for prediction based on multi-channel sequence data. Background Art

[0002] In the field of time series prediction, recurrent neural networks are widely used in finance, meteorology, and medicine because they can effectively capture dependencies in long sequence data. However, recurrent neural networks face the problem of insufficient feature extraction when dealing with complex multi-channel nonlinear and non-stationary data. In addition, recurrent neural networks are prone to the gradient vanishing problem when dealing with long time series data. Summary of the invention

[0003] The present application provides a method and related equipment for prediction based on multi-channel sequence data, aiming to solve the deficiencies or defects of the above-mentioned related technologies.

[0004] In a first aspect, the present application provides a method for prediction based on multi-channel sequence data, the method comprising:

[0005] Obtain multi-channel sequence data related to the prediction task specified by the user; the multi-channel sequence data includes state information of the prediction object associated with the prediction task at multiple time points in the past;

[0006] Perform multi-channel adaptive decomposition operation on multi-channel sequence data to obtain main frequency components and dynamic features in the multi-channel sequence data;

[0007] The main frequency components and dynamic features are input into the prediction model, and the output data of the prediction model is obtained as the prediction result. The prediction result includes the state information of the prediction object at multiple time points in the future.

[0008] In some embodiments, obtaining multi-channel sequence data related to a prediction task specified by a user includes:

[0009] Get the task content of the prediction task specified by the user;

[0010] Determine the prediction object associated with the prediction task based on the task content;

[0011] Obtain original multi-channel sequence data related to the prediction object;

[0012] The original multi-channel sequence data is cleaned and normalized to obtain multi-channel sequence data.

[0013] In some embodiments, performing a multi-channel adaptive decomposition operation on the multi-channel sequence data to obtain the main frequency components and dynamic features in the multi-channel sequence data includes:

[0014] Multi-channel adaptive Fourier decomposition is used to perform frequency decomposition on the time series data of each channel included in the multi-channel sequence data, and the common basis functions between multiple channels are aligned during the frequency decomposition to obtain the main frequency components and dynamic characteristics in the time series data of each channel.

[0015] In some embodiments, the training process of the prediction model includes:

[0016] Combine long short-term memory networks and recurrent neural networks to build an initial prediction model;

[0017] Obtaining training samples corresponding to the prediction object; the training samples include historical main frequency components and historical dynamic features extracted from historical multi-channel sequence data corresponding to the prediction object;

[0018] The initial prediction model is trained using the training samples until the preset end condition is met to obtain a trained prediction model.

[0019] In some embodiments, the prediction object is an electrocardiogram signal waveform; the multi-channel sequence data includes electrocardiogram signals of multiple leads within a specified period of time in the past; the main frequency component is information used to reflect the periodic changes of the electrocardiogram signal, and the dynamic feature is information used to reflect the short-term changes of the electrocardiogram signal; the state information of the prediction object at multiple time points in the future refers to the waveform of the electrocardiogram signal within a preset period of time in the future.

[0020] In some embodiments, the prediction object is the meteorological data of a city specified by the user; the multi-channel sequence data includes the daily meteorological data of the target city in the past multiple years, and the daily meteorological data includes index data of multiple meteorological indicators; the main frequency component is information used to reflect periodic changes in the meteorological data, and the dynamic feature is information used to reflect short-term changes in the meteorological data; the state information of the prediction object at multiple time points in the future refers to the daily weather conditions of the target city within a specified time period in the future.

[0021] In some embodiments, the prediction object is the stock price of multiple user-specified stocks; the multi-channel sequence data includes the stock data of each user-specified stock on each trading day in the past multiple years, and the stock data on each trading day includes the stock price and the trading volume; the main frequency component is the information used to reflect the periodic changes in the stock data, and the dynamic feature is the information used to reflect the short-term changes in the stock data; the state information of the prediction object at multiple time points in the future refers to the stock price and trading volume of each user-specified stock on each trading day within the specified time period.

[0022] In a second aspect, the present application provides a device for prediction based on multi-channel sequence data, the device comprising:

[0023] A data acquisition module is used to acquire multi-channel sequence data related to the prediction task specified by the user; the multi-channel sequence data includes state information of the prediction object associated with the prediction task at multiple time points in the past;

[0024] A decomposition module is used to perform a multi-channel adaptive decomposition operation on the multi-channel sequence data to obtain the main frequency components and dynamic features in the multi-channel sequence data;

[0025] The prediction module is used to input the main frequency components and dynamic features into the prediction model, and obtain the output data of the prediction model as the prediction result. The prediction result includes the state information of the prediction object at multiple time points in the future.

[0026] In a third aspect, the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the method for predicting based on multi-channel sequence data provided in the above embodiment is implemented.

[0027] In a fourth aspect, the present application provides a computer device comprising: one or more processors; a memory; and one or more computer programs, wherein the processor and the memory are connected via a bus, wherein the one or more computer programs are stored in the memory and are configured to be executed by the one or more processors, and characterized in that when the processor executes the computer program, the method for prediction based on multi-channel sequence data provided in the above embodiment is implemented.

[0028] The present application can execute a prediction task according to the user's instructions. When receiving the user's instructions, the application first obtains the multi-channel sequence data related to the prediction task specified by the user, and the multi-channel sequence data includes the state information of the prediction object associated with the prediction task at multiple time points in the past. Then, a multi-channel adaptive decomposition operation is performed on the multi-channel sequence data to obtain the main frequency components and dynamic features in the multi-channel sequence data; finally, the main frequency components and dynamic features are input into the prediction model, and the output data of the prediction model is used as the prediction result, and the prediction result includes the state information of the prediction object at multiple time points in the future.

[0029] Among them, this application provides higher quality features for the prediction model through multi-channel adaptive Fourier decomposition, which can improve the prediction accuracy. It also effectively reduces the dependence on long time series by combining multi-channel adaptive decomposition and long short-term memory network, thereby alleviating the gradient vanishing problem in long time series prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flowchart of a method for prediction based on multi-channel sequence data provided by an embodiment of the present application.

[0031] Figure 2It is a functional module block diagram of a device for performing prediction based on multi-channel sequence data provided in one embodiment of the present application.

[0032] Figure 3 It is a specific structural block diagram of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical solution and beneficial effects of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0034] In order to illustrate the technical solution described in this application, a specific embodiment is provided below for illustration.

[0035] See also Figure 1 , is a flowchart of a method for prediction based on multi-channel sequence data provided in an embodiment of the present application. This embodiment mainly illustrates the method by applying it to a computer device as an example.

[0036] The method for predicting based on multi-channel sequence data provided in an embodiment of the present application comprises the following steps:

[0037] S101, obtaining multi-channel sequence data related to a prediction task specified by a user; the multi-channel sequence data includes state information of a prediction object associated with the prediction task at multiple time points in the past.

[0038] The user sends a prediction instruction to the computer device to instruct the computer device to make a prediction. The user needs to specify the task content of the prediction task, which at least includes the prediction object.

[0039] In one embodiment, obtaining multi-channel sequence data related to a prediction task specified by a user includes: obtaining the task content of the prediction task specified by the user; determining a prediction object associated with the prediction task based on the task content; obtaining original multi-channel sequence data related to the prediction object; and performing data cleaning and normalization processing on the original multi-channel sequence data to obtain multi-channel sequence data.

[0040] The computer device can obtain multi-channel sequence data for realizing the prediction task based on the task content, and the multi-channel sequence data includes the status information of multiple channels corresponding to the prediction object at multiple time points in the past. The specific content of the multi-channel sequence data is related to the specific task content, and the specific data that needs to be obtained is different for different prediction tasks. For example, if the task content of the prediction task is to predict meteorological data, then the prediction object can be the meteorological data of a certain city, and the multi-channel time series data that needs to be obtained can be the index data of multiple meteorological indicators such as daily temperature, humidity, wind speed and precipitation of the city in the past multiple years; for another example, if the task content of the prediction task is electrocardiogram waveform prediction, then the prediction object can be the waveform of the electrocardiogram signal, and the multi-channel time series data that needs to be obtained includes the electrocardiogram signals of multiple leads within the past specified time period.

[0041] After obtaining the original multi-channel sequence data, it is cleaned and standardized to ensure the quality and consistency of the multi-channel sequence data used for subsequent processing, so that the subsequent multi-channel adaptive Fourier decomposition operation can extract higher quality features, thereby allowing the prediction operation to obtain more accurate prediction results.

[0042] Data cleaning can include missing data processing and outlier detection and processing. Normalization can use mean normalization, minimum-maximum normalization, or Z-score normalization to scale the data to a specific range, such as [0,1], so that the prediction model can be better generalized to new data.

[0043] S102, performing a multi-channel adaptive decomposition operation on the multi-channel sequence data to obtain main frequency components and dynamic features in the multi-channel sequence data.

[0044] The steps of performing a multi-channel adaptive decomposition operation on the multi-channel sequence data to obtain the main frequency components and dynamic features in the multi-channel sequence data specifically include: using multi-channel adaptive Fourier decomposition to perform frequency decomposition on the time series data of each channel included in the multi-channel sequence data, and aligning the common basis functions between multiple channels during the frequency decomposition to obtain the main frequency components and dynamic features in the time series data of each channel.

[0045] In this embodiment, Multiple Adaptive Fourier Decomposition (MFAD) uses adaptive basis functions to effectively capture the local characteristics of nonlinear and non-stationary signals, and in the frequency decomposition process, it mines the correlation between channels by aligning common basis functions between multiple channels. Based on the mined correlation between channels, the final prediction accuracy of the prediction model can be improved.

[0046] Multi-channel adaptive Fourier decomposition extends adaptive Fourier decomposition (AFD) to multi-channels. The adaptive Fourier decomposition itself will adaptively select the appropriate Fourier basis function, and then decompose the main frequency components and dynamic characteristics from the relevant time series data, and gradually extract the Fourier components by minimizing the reconstruction error until the error reaches a preset threshold.

[0047] The specific decomposition steps are as follows:

[0048] a. Initialize the parameters of the Fourier basis function, such as period and amplitude;

[0049] b. Use the least squares method to fit the time series data and obtain the initial Fourier component;

[0050] c. Dynamically adjust parameters and minimize reconstruction error through iterative optimization;

[0051] d. Extract the main Fourier components as input for subsequent recurrent neural network modeling.

[0052] Adaptive Fourier Decomposition Using Matching Pursuit Method to Identify Adaptive Basis Function Systems Each function B n It is called the modified Blaschke product. The modified Blaschke product of each basis function is defined as follows:

[0053]

[0054] Among them, a n is the complex coefficient in the unit disk D = {z∈C:|z|<1, and is an adaptive parameter chosen to maximize energy convergence at each decomposition level. For any sequence of a in the unit disk D n are orthogonal. The processed signal G(e jt ) can be expressed as the series expansion form of the adaptive basis function:

[0055]

[0056] Among them, A n For nth th These coefficients are obtained by combining the signal G and the basis function B n The inner product of <G,B n >Calculated.

[0057] In multi-channel adaptive Fourier decomposition, all channels use the same number of single components and the same basis functions to maintain the alignment of common oscillations among channels. The total energy to be maximized in multi-channel adaptive Fourier decomposition is:

[0058]

[0059] Where C is the total number of channels, G c,n represents the processed signal of the cth channel in the nth decomposition level. This formula is used to determine the optimal parameter a that maximizes the energy of all channels n .

[0060] Multi-channel adaptive Fourier decomposition can adaptively extract the main frequency components and dynamic features in the time series data of each channel, making the features input to the LSTM-RNN model highly representative. In addition, the features extracted by multi-channel adaptive Fourier decomposition can capture the key information in the data, enabling LSTM-RNN to more accurately model complex time series data, thereby significantly improving the accuracy of the prediction results.

[0061] Furthermore, multi-channel adaptive Fourier decomposition can adaptively select the decomposition basis function according to the characteristics of the relevant time series data, avoiding the limitations of traditional fixed basis functions in processing nonlinear and non-stationary signals. This adaptive capability can make the final extracted features more representative, which will further improve the predictive performance of the prediction model.

[0062] S103: Input the main frequency components and dynamic features into the prediction model, and obtain the output data of the prediction model as the prediction result, where the prediction result includes the state information of the prediction object at multiple time points in the future.

[0063] The prediction model of this embodiment is trained based on LSTM-RNN (i.e., long short-term memory network-recurrent neural network). Among them, the LSTM-RNN model can effectively capture long-term and short-term dependencies. On this basis, combined with multi-channel adaptive Fourier decomposition, the prediction model can further enhance its ability to capture dynamic changes in time series data and improve the overall prediction performance. In addition, on the basis of the recurrent neural network, combined with multi-channel adaptive Fourier decomposition and long short-term memory network, it can effectively reduce the dependence on long time series, alleviate the gradient vanishing problem in long sequence prediction, thereby improving the training speed and reasoning efficiency of the model.

[0064] In addition, the combination of feature extraction operations based on multi-channel adaptive Fourier decomposition and sequence modeling of LSTM-RNN can enable the prediction model to exhibit high generalization ability on different data sets. This makes the prediction method provided in this application not only suitable for single fields such as financial market prediction, weather forecast, and medical data analysis, but can also be extended to other application scenarios that require processing complex multi-channel time series data.

[0065] Regarding the training process of the prediction model, in one embodiment, the training process of the prediction model includes:

[0066] a. Combine long short-term memory network and recurrent neural network to build the initial prediction model;

[0067] b. Obtaining training samples corresponding to the predicted object; the training samples include historical main frequency components and historical dynamic features extracted from historical multi-channel sequence data corresponding to the predicted object;

[0068] c. Use the training samples to train the initial prediction model until the preset end conditions are met to obtain a trained prediction model.

[0069] When training the prediction model, this embodiment uses the main frequency components and dynamic features extracted by multi-channel adaptive Fourier decomposition as the input of the prediction model to train the prediction LSTM-RNN model. Through its cyclic structure, LSTM-RNN can capture the short-term and long-term dependencies in time series data, and then accurately predict the data at future time points. In addition, the LSTM unit can effectively alleviate the gradient vanishing problem through its memory ability, so that the prediction model can maintain stable performance when processing long sequence data.

[0070] During the model training process, historical data is used to supervise the LSTM-RNN model and optimize the model parameters. Backpropagation algorithms (such as BPTT, Backpropagation Through Time) and gradient descent algorithms are used to ensure the convergence and stability of the model. At the same time, through methods such as cross-validation, the model's hyperparameters (such as learning rate, number of hidden layers, etc.) are optimized to prevent overfitting and improve the generalization ability of the model.

[0071] Furthermore, in order to cope with the dynamic changes of data, the present application also provides a mechanism for dynamically updating the prediction model, that is, after the prediction model is trained, the prediction model is predicted based on new multi-channel sequence data related to the prediction object, and the model parameters are adjusted based on the prediction results. This mechanism allows the multi-channel adaptive Fourier decomposition and prediction model to be continuously updated and self-adjusted to ensure the prediction accuracy of the model in different time periods.

[0072] The prediction method provided in the above embodiment will be described in detail below using multiple prediction tasks with different task contents as examples.

[0073] In one embodiment, the prediction object is an electrocardiogram signal waveform; the multi-channel sequence data that needs to be acquired at this time includes electrocardiogram signals of multiple leads within a specified period of time in the past; the main frequency components extracted by the feature extraction operation are information used to reflect the periodic changes of the electrocardiogram signal, and the dynamic features are information used to reflect the short-term changes of the electrocardiogram signal; accordingly, the state information of the prediction object at multiple time points in the future included in the prediction result refers to the waveform of the electrocardiogram signal within a preset period of time in the future.

[0074] In another embodiment, the prediction object is the meteorological data of a city specified by the user. In this case, the multi-channel sequence data that needs to be acquired includes the daily meteorological data of the target city in the past multiple years (including index data of multiple meteorological indicators). The main frequency components extracted by the feature extraction operation are information used to reflect periodic changes in the meteorological data, and the dynamic features are information used to reflect short-term changes in the meteorological data. Accordingly, the state information of the prediction object at multiple time points in the future included in the prediction results refers to the daily weather conditions of the target city within a specified time period in the future.

[0075] In another embodiment, the prediction object is the stock price of multiple user-specified stocks. In this case, the multi-channel sequence data to be acquired includes the stock data (including stock price and trading volume) of each user-specified stock on each trading day in the past multiple years. The main frequency components extracted by the feature extraction operation are information used to reflect periodic changes in the stock data, and the dynamic features are information used to reflect short-term changes in the stock data. Accordingly, the state information of the prediction object at multiple time points in the future included in the prediction results refers to the stock price and trading volume of each user-specified stock on each trading day within the specified time interval.

[0076] The present application also provides a device for prediction based on multi-channel sequence data. In some embodiments, see Figure 2 , the device comprises:

[0077] The data acquisition module 10 is used to acquire multi-channel sequence data related to the prediction task specified by the user; the multi-channel sequence data includes state information of the prediction object associated with the prediction task at multiple time points in the past;

[0078] A decomposition module 20, used to perform a multi-channel adaptive decomposition operation on the multi-channel sequence data to obtain the main frequency components and dynamic features in the multi-channel sequence data;

[0079] The prediction module 30 is used to input the main frequency components and dynamic features into the prediction model, and obtain the output data of the prediction model as the prediction result. The prediction result includes the state information of the prediction object at multiple time points in the future.

[0080] In some embodiments, the data acquisition module 10 includes:

[0081] A task content acquisition unit, used to acquire the task content of the prediction task specified by the user;

[0082] A prediction object determination unit, used to determine a prediction object associated with a prediction task based on the task content;

[0083] An original data acquisition unit, used to acquire original multi-channel sequence data related to the prediction object;

[0084] The preprocessing unit is used to perform data cleaning and normalization processing on the original multi-channel sequence data to obtain multi-channel sequence data.

[0085] In some embodiments, the decomposition module 20 includes:

[0086] The frequency decomposition unit is used to perform frequency decomposition on the time series data of each channel included in the multi-channel sequence data by using multi-channel adaptive Fourier decomposition, and align the common basis functions between multiple channels during frequency decomposition to obtain the main frequency components and dynamic characteristics in the time series data of each channel.

[0087] In some embodiments, the apparatus further comprises a prediction model training module, the module comprising:

[0088] An initial model building unit, used to build an initial prediction model by combining a long short-term memory network and a recurrent neural network;

[0089] A training sample acquisition unit is used to acquire training samples corresponding to the prediction object; the training samples include historical main frequency components and historical dynamic features extracted from historical multi-channel sequence data corresponding to the prediction object;

[0090] The training unit is used to train the initial prediction model using the training samples until a preset end condition is met to obtain a trained prediction model.

[0091] In some embodiments, the prediction object is an electrocardiogram signal waveform; the multi-channel sequence data includes electrocardiogram signals of multiple leads within a specified period of time in the past; the main frequency component is information used to reflect the periodic changes of the electrocardiogram signal, and the dynamic feature is information used to reflect the short-term changes of the electrocardiogram signal; the state information of the prediction object at multiple time points in the future refers to the waveform of the electrocardiogram signal within a preset period of time in the future.

[0092] In some embodiments, the prediction object is the meteorological data of a city specified by the user; the multi-channel sequence data includes the daily meteorological data of the target city in the past multiple years, and the daily meteorological data includes index data of multiple meteorological indicators; the main frequency component is information used to reflect periodic changes in the meteorological data, and the dynamic feature is information used to reflect short-term changes in the meteorological data; the state information of the prediction object at multiple time points in the future refers to the daily weather conditions of the target city within a specified time period in the future.

[0093] In some embodiments, the prediction object is the stock price of multiple user-specified stocks; the multi-channel sequence data includes the stock data of each user-specified stock on each trading day in the past multiple years, and the stock data on each trading day includes the stock price and the trading volume; the main frequency component is the information used to reflect the periodic changes in the stock data, and the dynamic feature is the information used to reflect the short-term changes in the stock data; the state information of the prediction object at multiple time points in the future refers to the stock price and trading volume of each user-specified stock on each trading day within the specified time period.

[0094] The device for predicting based on multi-channel sequence data provided in one embodiment of the present application and the method for predicting based on multi-channel sequence data provided in the present application belong to the same concept. The specific implementation process is detailed in the full text of the specification and will not be repeated here.

[0095] An embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the method for predicting based on multi-channel sequence data provided in an embodiment of the present application is implemented.

[0096] Figure 3 A specific structural block diagram of a computer device provided in an embodiment of the present application is shown. The computer device 100 includes: one or more processors 101, a memory 102, and one or more computer programs, wherein the processor 101 and the memory 102 are connected via a bus, the one or more computer programs are stored in the memory 102, and are configured to be executed by the one or more processors 101, and when the processor 101 executes the computer program, the method for predicting based on multi-channel sequence data provided in an embodiment of the present application is implemented.

[0097] It should be understood that each step in each embodiment of the present application is not necessarily performed in sequence according to the order indicated by the step number. Unless there is clear explanation in this article, the execution of these steps does not have strict order restriction, and these steps can be performed in other orders. Moreover, at least a part of the steps in each embodiment can include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a part of other steps or sub-steps or stages of other steps.

[0098] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink), DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0099] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0100] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for prediction based on multi-channel sequence data, characterized in that: The method comprises: Acquire multi-channel sequence data related to a prediction task specified by a user; the multi-channel sequence data includes state information of a prediction object associated with the prediction task at multiple time points in the past; Performing a multi-channel adaptive decomposition operation on the multi-channel sequence data to obtain main frequency components and dynamic features in the multi-channel sequence data; The main frequency components and the dynamic features are input into a prediction model, and output data of the prediction model is obtained as a prediction result, wherein the prediction result includes state information of the prediction object at multiple time points in the future.

2. The method according to claim 1, characterized in that Get multi-channel sequence data related to the prediction task specified by the user, including: Get the task content of the prediction task specified by the user; Determining a prediction object associated with the prediction task based on the task content; Acquire original multi-channel sequence data related to the prediction object; The original multi-channel sequence data is cleaned and normalized to obtain the multi-channel sequence data.

3. The method according to claim 1, characterized in that Performing a multi-channel adaptive decomposition operation on the multi-channel sequence data to obtain main frequency components and dynamic features in the multi-channel sequence data includes: Multi-channel adaptive Fourier decomposition is used to perform frequency decomposition on the time series data of each channel included in the multi-channel sequence data, and common basis functions between multiple channels are aligned during frequency decomposition to obtain the main frequency components and dynamic characteristics in the time series data of each channel.

4. The method according to claim 1, characterized in that The training process of the prediction model includes: Combine long short-term memory networks and recurrent neural networks to build an initial prediction model; Acquire a training sample corresponding to the prediction object; the training sample includes historical main frequency components and historical dynamic features extracted from historical multi-channel sequence data corresponding to the prediction object; The initial prediction model is trained using the training samples until a preset end condition is met to obtain a trained prediction model.

5. The method according to claim 1, characterized in that The prediction object is an electrocardiogram signal waveform; The multi-channel sequence data includes electrocardiogram signals of multiple leads within a specified time period in the past; The main frequency component is information for reflecting the periodic changes of the electrocardiogram signal, and the dynamic feature is information for reflecting the short-term changes of the electrocardiogram signal; The state information of the predicted object at multiple time points in the future refers to the waveform of the electrocardiogram signal within a preset time period in the future.

6. The method according to claim 1, characterized in that The prediction object is the meteorological data of the city specified by the user; The multi-channel sequence data includes daily meteorological data of the target city in the past multiple years, and the daily meteorological data includes index data of multiple meteorological indicators; The main frequency component is information used to reflect periodic changes in meteorological data, and the dynamic feature is information used to reflect short-term changes in meteorological data; The state information of the forecast object at multiple time points in the future refers to the daily weather conditions of the target city within a specified time interval in the future.

7. The method according to claim 1, characterized in that The prediction object is the stock price of multiple user-specified stocks; The multi-channel sequence data includes the stock data of each user-specified stock on each trading day in the past multiple years, and the stock data on each trading day includes the stock price and trading volume; The main frequency component is information used to reflect periodic changes in stock data, and the dynamic feature is information used to reflect short-term changes in stock data; The state information of the prediction object at multiple time points in the future refers to the stock price and trading volume of each user-specified stock on each trading day within a specified time interval.

8. A device for prediction based on multi-channel sequence data, characterized in that: The device comprises: A data acquisition module, used to acquire multi-channel sequence data related to a prediction task specified by a user; the multi-channel sequence data includes state information of a prediction object associated with the prediction task at multiple time points in the past; A decomposition module, used for performing a multi-channel adaptive decomposition operation on the multi-channel sequence data to obtain main frequency components and dynamic features in the multi-channel sequence data; The prediction module is used to input the main frequency components and the dynamic features into a prediction model, and obtain the output data of the prediction model as a prediction result, wherein the prediction result includes the state information of the prediction object at multiple time points in the future.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method for prediction based on multi-channel sequence data as described in any one of claims 1 to 7 are implemented.

10. A computer device, characterized in that: include: one or more processors; Memory; as well as One or more computer programs, the processor and the memory are connected via a bus, wherein the one or more computer programs are stored in the memory and are configured to be executed by the one or more processors, wherein the processor implements the steps of the method for predicting based on multi-channel sequence data as described in any one of claims 1 to 7 when executing the computer program.

Images