| With the rapid development of China’s computer,Internet and big data technologies and the vigorous development of the securities market such as stock market,the programmatic intelligent investment model is rapidly emerging and developing.The development of artificial intelligence technology has greatly improved the ability of machine learning.Computers quickly process large amounts of data,summarize valuable information and rules,develop new trading strategies,and modify and refine existing investment models,all of which are faster and more efficient.This paper mainly studies the establishment of a programmatic investment model applied in the stock market through algorithms and physical models.The data mining method and the machine learning algorithm are used to classify and screen the factors affecting the stock price,and then select the market factors that have a significant impact on the stock price.Considering the actual situation of the stock market and the impact of domestic macroeconomic conditions on stock prices,the classical physical model Kalman filter model is introduced to improve the effective factors of the stock market,mainly state space model,innovation process and Kalman gain calculation.In order to make the stock market effective factor can more significantly reflect the trend of stock prices.The multi-factor model is used to program stocks from the fundamentals and technology to form a stock pool.Through the decision tree algorithm,the composite returns,excess returns and correlations of the fundamental candidate factors are analyzed.Through the logistic regression algorithm,the predicted probability of the technical candidate factors is analyzed to establish an advantageous stock investment object.And build investment ideas and risk control methods for investment models.Finally,the programmaticization of the investment strategy model is realized.In empirical research,the programmatic investment model is used to back up the historical data of the stock pool,and the programmatic investment model is used for testing in real-time simulation,and the programmatic investment model is improved from the aspects of net rate of return and maximum retracement.This paper studies and establishes a procedural investment model applied in the stock market through machine learning algorithm and physical model.Through historical data and real-time simulation,the procedural investment model is continuously improved in order to obtain excess return through probability advantage and help investors get higher returns in the stock market. |
