The Quantitative Machinery Decision-Making Strategy Model Based on Neural Network
Yifei Hao, Qian Zhang, Bo Wang, Shenglin Xu · IOP Conference Series Materials Science and Engineering · 2019
Abstract High-frequency data and quantitative models are the hot issues in the machinery field. In order to build a quantitative strategy with steady returns and relatively low risks based on high frequency data, this paper establishes a quantitative machinery selection strategy based on difference flow and a time-selection strategy based on neural network. Through our research on China’s machinery market, we find that when using fund flow intensity as an indicator, the market yield of the portfolio with large capital flow is relatively high. After selecting some companies with potential upside, we used a neural network algorithm to predict the value between both sides. Using these technical indicators and historical information, we use neural networks to predict the next day’s closing value. Finally, we simulated the decision with Matlab, and analyzed the profitability and risk factors of our strategy in different periods of the machinery market, and drew the final conclusion.