The Research Of Multi-Factors Strategies Based On Decision Tree Model

The quantitative investment analysis field,which has been booming in the western countries,is on the way in Chinese stock market,and in recent years it has become a new force equivalently compared with Technical analysis and Fundamental analysis in A stock market.The quantitative investment is a methology which makes investment strategies happen by choosing the appropriate models,parameters and indicators as a whole through computer technology.The quantitative investment technology makes the breadth and depth of the investor's investment being greatly expanded.Especially with the help of machine learning algorithm and cloud computing,it will be easier and quicker for investor to mine the underlying valuable information through the massive financial data.In this paper,we apply the traditional Multi-Factor Model and four different Decision tree classification algorithms to evaluate the various financial indicators and construct a well-being portfolio and gaining extra profits.This thesis mainly includes the following contents:The first chapter expounds the background significance and research status,then it describes the framework of the problem;The second chapter focuses on the preliminary knowledge of Multi-Factor model and machine learning algorithm and discuss the Technical architecture of implementing the Multi-Factor model as well;The third chapter introduces four different categories of fundamental indicators which are underlying profitable and quantifiable.And based on real data from wind Database,testing their effectiveness on profit,finding the outstanding profitable indicators.The forth chapter,based on the outstanding profitable factors,apply the multi-factor model with decision tree algorithm and real data from wind Database,get the final quantitative strategies about stock portfolio and its back testing results.The fifth chapter is the summary of this paper in order to focus on improving factors.