Forecasting Stock Price Using Partial Least Squares Regression

Financial markets are systems where buyers and sellers interact and various transactions of ownership claims (stocks) takes place. These transactions are made with the base purpose of earning profits, but the biggest problem with these transactions is that the profit depends on the rise in the prices of the stock and one may suffer loss if the price goes below the price at which the stock was initially purchased. The objective of preparing this model is to treat the market as a mathematical problem and try and use statistical algorithms to make the behavior of the market somewhat predictable and since data mining tools can be very effective in predicting future trends and preferences allowing decision makers to make forecasting on the data from the past. This study describes a model built for predicting the behavior of certain stocks based on their behavior in past.