Contrast principal component data analysis method based on deep neural network

Abstract

According to the comparison principal component data analysis method based on the deep neural network, nonlinear mapping can be flexibly performed on data through training, and then comparison information existing in the data is extracted. Comprising the steps of performing mean centralization on original data vectors of which comparison information needs to be extracted, randomly selecting a batch of data from training data, inputting the data into a neural network for forward propagation, and mapping the data to a new feature space through a target network and a background network; performing mean centralization operation on the output of the target network and the background network again, calculating a sample covariance matrix, and performing comparative principal component analysis to obtain a comparative projection direction matrix; and calculating a gradient value of the loss function for network output, iterating the weights of the target network and the background network through a back propagation algorithm, performing forward propagation and comparative principal component analysis on all data, and projecting the network output in a comparative projection direction to obtain a comparative principal component.

Description

technical field [0001] The invention relates to a comparative principal component data analysis method based on a deep neural network, belonging to the field of data analysis, in particular to high-dimensional nonlinear data analysis technology generated in the fields of intelligent manufacturing, biomedicine and quantitative finance. Background technique [0002] With the continuous development of data acquisition and storage technology, a large amount of high-dimensional complex data continues to emerge in the process of intelligent manufacturing. Although high-dimensional data contains rich information, the high dimensionality also brings challenges to data analysis. Data dimensionality reduction and feature extraction technology can reduce the dimension of data and reduce the influence of redundant features in the data, so it occupies an important position in the process of fusion analysis of high-dimensional data. [0003] Principal Component Analysis (PCA) is a classic...

AI Technical Summary

Problems solved by technology

So far, there is no method that can effectively improve the ability to extract contrast information in real-world nonlinear data

Images