Data Analysis of Portfolio Optimization using Artificial Neural Network in China’s Stock Market*

This paper presents a neural network model for studying the portfolio optimization in China’s stock market. An investment scheme is proposed by using an artificial intelligence algorithm based on the trained neural network. In the empirical study, six key indicators of historical transaction data for stocks are used as the input characteristics of our neural network model, which are the opening price, the closing price, the trading volume, the highest and lowest price on the day, and Week time scale. Besides, we use Python technology to process the collected stock data and develop the learning process of the studied neural network for predicting the closing price for the next period. Based on the presented data analysis in the paper, the investors of the stock market in China could make a rational decision-making on the portfolio optimization.