Stock Market Forecasting Using Hidden Markov Model

Forecasting stock price or financial markets has been one of the biggest challenges to the AI community. A Hidden Markov Model (HMM) is a finite state machine which has some fixed number of states. It provides a probabilistic framework for modelling a time series of multivariate observations. Hidden Markov models were introduced in the beginning of the 1970’s as a tool in speech recognition. This model based on statistical methods has become increasingly popular in the last several years due to its strong mathematical structure and theoretical basis for use in a wide range of applications.
This poject use the well-established Hidden Markov Model (HMM) technique to forecast stock price for some of the airlines. It used given datasets (four Corporations) and implement method in the Stock Market Forecasting Using Hidden Markov Model: A New Approach paper. The dataset files (Corporation1-4) contain daily stock prices (almost 2500 working days) for 4 different corporations in order- Close, Open, High, Low. (Features = Close price, Open price, High price, Low price) attention that, the data at the end of row datasets is related to the first days, the first rows are related to the recent days.
We kept aside the recent 200 observations for testing (the first 200 rows of each dataset) and used rest of the observations for training the model. We predict the prices for the past 200 days, starting from the 200th day (200 recent days) separated for each feature for all datasets and then plot actual price (test set) and predicted price (for each feature separated - horizontal axes=200 days, vertical axis = price) and report the MAPE for each dataset.