Two-Factor Fuzzy Hidden Markov Model For Stock Price Index Forecasting

• Main Authors: WEN,CHUN-YI, 溫鈞怡
• Other Authors: WANG,HSIEN-LUN
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/88881689512151945575

碩士 === 明新科技大學 === 管理研究所碩士在職專班 === 103 === An effective forecasting system can help companies or individuals to reduce investment risk. Existing research has proposed numerous ways to predict, and most of them strive for higher precision of prediction. In recent years, the hidden Markov model began to be applied to social science research, and proved to have excellent performance in predicting returns. To assess their empirical advantage, this study uses two-factor Hidden Markov Model to predict four stock indexes of daily trading information, including : Taiwan weighted stock index (for the period 1990-1999), the Nasdaq stock index (for the period 1990-1999), standard and poor's 500-stock index (for the period 2000-2009), and the Dow Jones industrial stock index (for the period 2000-2009). The empirical results show that our predictive models have better prediction performance of lower forecast errors than do many other existing models.