Time series forecasting method and equipment and system adopting same

Abstract

The embodiment of the invention discloses a time series forecasting method and equipment and a system adopting the same. The method comprises the following steps: training acquired time series data to obtain a training data set; utilizing the training data set to train a selected predictor group to generate a diversified predictor group; extracting sparse signals to restructure majorized functions and solving the weighting factor of the diversified predictor group; and intercepting predictors with non-zero weighting factors to forecast the time series data. In the embodiment of the invention, the weighting factor of the diversified predictor group serves as the sparse signals for restructuring and solving the corresponding majorized functions, and as the obtained weighting factor has sparseness, the obtained weighting factor is utilized to achieve time series data forecasting and checking for the predictors with non-zero weighting factors; and as the quantity of the predictor group is reduced, the checking is sped up and the forecasting accuracy is improved.

Description

technical field [0001] The present invention relates to the technical field of data prediction, and more specifically, to a time series prediction method, device and method. Background technique [0002] The time series forecasting method is based on the ordered observation data sets associated with the time sequence, and uses the stochastic process theory and mathematical statistics methods to study the statistical laws obeyed by the data sets, so as to infer the data development trend and guide the solution of practical problems. The time series prediction method has been widely applied to fields such as industry, address, ecology, economy, meteorology, and medicine. [0003] In the prior art, the most used tools for time series forecasting are predictors such as neural networks, decision trees and support vector machine methods based on machine learning algorithms. In order to achieve better time series forecasting results, multiple identical or Different predictors are ...

AI Technical Summary

Problems solved by technology

[0005] Since the output results of multiple predictors are fused, the average weighting method, the least squares weighting method, and the linear programming coefficient weighting method are used. The average weighting method and the least squares weighting method are all predictors. When verifying the predictor data, it is necessary to verify the predictors one by one, which undoubtedly makes the verification process slow. Although the linear programming coefficient weighting method is relatively fast in the verification process, the accuracy of time series prediction is not ideal

Images