Daily Trading Decisions Based on Prediction and Optimization

Abstract

With the continuous development of the social economy, people's disposable income continues to increase, and more and more families begin to enter the ranks of financial investment. In order to seek greater investment returns, it is necessary to formulate the best trading strategy. In this paper, BP neural network is used for price forecasting, and according to the principles of controlling risks and increasing returns, a daily transaction decision-making model based on forecasting, optimization and risk control is established to determine the optimal investment portfolio. And through the volatility assessment, the final decision on whether to buy or sell gold and bitcoin. At the same time, this paper establishes a trading strategy evaluation model based on Analytic Hierarchy Process (AHP) to evaluate three trading strategies. The analysis shows that the trading strategy evaluation model can serve as the basis for optimal strategy evaluation.