Program Trading

• Main Authors: CHEN, SHENG-WEN, 陳聖文
• Other Authors: HUNG, MING-CHIN
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/7f9d6t

碩士 === 東吳大學 === 財務工程與精算數學系 === 106 === Recently, data science grows rapidly, the algorithms of data mining and machine learning are continuously advanced and programmable. Prediction is one of the most popular topic in recently years. This thesis base on the easy-to-get finance data and various stock indexes to build a predict system by R language. The open source R language and its libraries can simulate the automatic trading by using algorithms through packages. This paper shows predict results of the Artificial Neural Network, support vector machine and MARS three algorithms. According to the literatures, program trading can get good rate of return by these three algorithms in Taiwan Futures Market. The result shows that the support vector machine can get better rate of return in three algorithms. Even the 2008, Financial crisis, support vector machine still can get positive return rate. Hope this thesis can help people who have an interest in investing