Dimensionality reduction method, device, computer equipment and storage medium for time series data

Abstract

The present application relates to a dimensionality reduction method, device, computer equipment and storage medium for time series data. The method comprises: inputting the time series data to be processed into a pre-trained dimensionality reduction neural network model in chronological order, obtaining the hidden layer vector at the last moment in the dimensionality reduction neural network model; based on the t-SNE dimensionality reduction algorithm , performing dimensionality reduction processing on the hidden layer vectors to obtain a dimensionality reduction data set; the dimensionality reduction data set is a set of hidden layer vectors reduced to two dimensions. Using this method can realize the dimensionality reduction processing of time series data in matrix form.

Description

technical field [0001] The present application relates to the technical field of data analysis, in particular to a method, device, computer equipment and storage medium for dimensionality reduction of time series data. Background technique [0002] In the field of machine learning, a sample often contains dozens or even hundreds of features, and each feature is a dimension. In order to visualize such high-dimensional data, it is necessary to reduce the feature dimension of these data to two or three dimensions. displayed on the computer. [0003] Current data dimensionality reduction techniques can only be used for data that only occupies one row of a sample, but for sequence-type human data. For example, using the sensor data collected by mobile phones to predict the actions of mobile phone users, an action consists of multiple time steps, which is represented by multiple lines in the data set, each line represents a time step, and the features of each step include x, Acc...

AI Technical Summary

Problems solved by technology

For this type of time series data, traditional dimensionality reduction methods cannot perform dimensionality reduction processing

That is, the traditional data dimensionality reduction method can only reduce the dimensionality of high-dimensional vectors to low-dimensional vectors, and cannot reduce the dimensionality of time series data in the form of matrices.

Images