Investment portfolio optimization using evolutionary strategies
Talking about investment, people may simply think to put money in a saving account in a bank. However, such small amount earned by bank interest cannot keep ahead with the growing inflation rate and the value of the money is depreciated. Therefore, more and more people are interested in putting a po...
Saved in:
| 主要作者: | |
|---|---|
| 其他作者: | |
| 格式: | Final Year Project |
| 語言: | English |
| 出版: |
2013
|
| 主題: | |
| 在線閱讀: | http://hdl.handle.net/10356/53412 |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | Talking about investment, people may simply think to put money in a saving account in a bank. However, such small amount earned by bank interest cannot keep ahead with the growing inflation rate and the value of the money is depreciated. Therefore, more and more people are interested in putting a portion of their savings in more aggressive investments, such as stock, which can offer a handsome return that wins inflation and create more wealth.
By such, a stock forecasting model with a higher precision which produces theoretical significance and applicable value is highly preferred for investors’ reference. In this project, the feasibility of stock trend predication by using genetic algorithm of neural network is discussed. Neural network is to calculate the weights and valve value of the historical stock data by learning it with certain mean-square-error (MSE) [1] and give the predicated result with certain confidence level, which will be discussed in details in this report.
In this project, the simulation is done based on BP Algorithm [2] and by conducting through MATLAB. Google Inc. (GOOG) stock will be the case-study to apply the trained network and a fair-good effect has been achieved.
KEYWORDS Stock Predication, Genetic Algorithm, Neural Network, Back Propagation, Matlab |
|---|