Using K-means and Support Vector Machine For Stock Price Trends Prediction
碩士 === 元智大學 === 資訊管理學系 === 101 === Besides serving as an indicator of economic sentiment, movement of stock market is also a leading indicator of future economic growth. Taiwan's stock market fluctuation is also closely coupled with world's major stock markets. This study hopes to predict...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Online Access: | http://ndltd.ncl.edu.tw/handle/78773251029948861614 |
| Summary: | 碩士 === 元智大學 === 資訊管理學系 === 101 === Besides serving as an indicator of economic sentiment, movement of stock market is also a leading indicator of future economic growth. Taiwan's stock market fluctuation is also closely coupled with world's major stock markets. This study hopes to predict stock price movement effectively and to examine the coupling effect.
Share prices and relevant technical indicators were used as inputs of stepwise regression analysis to find out the influential factors. The results were then subjected to k-means clustering analysis to improve predictive ability.
The second phase of this study used support vector machine (SVM) to build a model to predict future share price movement and to investigate whether world’s stock market indexes affect Taiwan’s share price movement. The SVM model’s performance was also compared with another model which was based on back propagation neural network.
|
|---|