Author(s)

Peng Liang, Jie Yang, Jian Zhou, Zongyuan Tan, Zhenpei Shan

Corresponding Author

Peng Liang

Abstract

This paper aims to propose a method called linear Alexnet to analyze the fluctuation of stock price. In this study, the fluctuation of stock price can be sumarized as three main factors. Based on the simulation results, we found that deep learning network is suitable to simulating and analyzing the current economic models, in particular certain fluctuation models. Of course, to further boost the performance of training, we add certain Gauss noises for avoiding the generalization of dataset.

Keywords

Stock fluctuation, Linear alexnet, Deep learning, PCA