Forecasting in Stock Market is a complex and important problem. The market trend keeps on changing every now and then. Forecasting has posed many challenges for the researchers working in the area of financial mathematics. The market is influenced by many direct and indirect factors and this has made financial models more Vern able. The researchers have proposed many models,8,9,7,16. The results are good under certain circumstances but still not full proof. In this paper the authors have tried to use Hidden Markov Model (HMM)17. HMM is widely used in the field of pattern recognition like Gesture recognition, Texture recognition, handwriting, bioinformatics etc. Here the HMM is trained by using Baum-Welch algorithm11 and then likelihood values are found from the past data sets. Based on likelihood values future stock price is predicted. The proposed model was implemented using MATLAB tool box. Suggested technique provides the more accurate results than other techniques.
Hidden Markov Model, stock market, prediction, likelihood value, Baum-Welch algorithm
