| Summary: | 碩士 === 輔仁大學 === 資訊管理學系碩士在職專班 === 106 === This study is mainly to explore the prediction of the decision tree algorithm for the Taiwan stock futures, and with the fund management model as a trading strategy, and perform performance backtest experiments. This study uses PowerLanguage to write programs, automatically collect historical data on the MultiCharts trading platform, as CART and random forest algorithm training materials, predict through the C++ programming method, return the prediction results to MultiCharts, and write with PowerLanguage. The fund management module program builds functions and signals in MultiCharts and performs performance backtesting experiments with MultiCharts Portfolio traders.
The results of this study show that the algorithm predicts that the trading strategy of the matching fund management module is better than the performance of the general single strategy, and the experimental results confirm the investment of Taiwan stock futures with random forest algorithm and fixed ratio fund management mode. Trading strategy, the highest return on investment for trading results; small Taiwan index futures with random forest calculus forecast, and Kelly formula, fixed fraction method, fixed ratio method, best F value, Larry Willams, investment of five fund management modules The trading strategy, the results can be stable and profitable.
|
|---|
