Predict price of a stock 3

Brzoska, Karl

Student Work

Predict price of a stock 3

Public Deposited

This IQP looked at different methods to predict stock prices. 20 stocks were chosen from a sector in the stock market to make short term predictions. The first model used was a simple linear model, which did not prove to have a good prediction rate. After adding a fourier series, the prediction accuracy increased. These predictions stayed within the error bounds with an average of 0.6 days. Error bounds, which were created by taking the average deviation between the actual stock price and the predicted stock price from the linear model, were used to gauge how accurate our predictions were. If the predicted stock price stayed within the error bounds, the prediction was determined accurate. When taking into account the market influence, the prediction accuracy increased even further with predictions staying within the error bounds with an average of 3 days. The final method used was a Savitsky Golay method, which was inconsistent and no conclusions could be drawn.

This report represents the work of one or more WPI undergraduate students submitted to the faculty as evidence of completion of a degree requirement. WPI routinely publishes these reports on its website without editorial or peer review.

Creator