Portfolio Optimization Using Machine Learning and Deep Learning methods to predict rate of return. Then use the predicted date to build portfolio optimization models.
- Tradtional Mean-Variance model use mean return as rate of return, but that's not suitable for short term investment;
- Nomal distuibution hypothesis not always works in reality;
- Variance to stand the risk of investment is too roughly;
- First use data of hs300 to test this method;
- Then more data of stocks will be used;
- First use linear regrssion;
- Then use xgboost and so on;
- Add the views of investors to the model;
- Others