Quantitative Trading Model of Two Shares Based on NARX Neural Network and Dynamic Programming

Abstract

This paper focuses on the quantitative trading of two shares based on historical data. In modern financial research, quantitative transaction is an important research topic. With the rapid development of the financial market, it is of great practical significance to accurately predict the future data through historical data and make reasonable decisions. This paper aims to help traders improve their decision-making ability and optimize asset allocation by building a quantitative trading decision-making model. This paper analyzes the advantages and disadvantages of the model, and proves that the established model has high robustness, accuracy and popularization significance. According to the experimental results and research conclusions, the memorandum provides information and suggestions for investors' investment choices.