Development of Unscented Kalman Filter Algorithm for stock price estimation

Abstract

Stock market is established in order to bring together the stock sellers and buyers. Securities often traded in the stock market are shares. Shares are securities as proof of participation or ownership of a person or legal entity in a company. In choosing a safe and appropriate investment in stocks, investors need a way to assess the price of the shares to be purchased or the ability of the stock to provide dividends in the future, so as to optimize profits. The correct way to analyze the risk for investors in investing is to estimate the stock price. The purpose of this paper is to analyze the comparison of share price estimates using the Unscented Kalman Filter (UKF) and Unscented Kalman Filter Square Root (UKF-SR) methods. The simulation results show that both methods have a significantly high accuracy of less than 2%. We conclude that the two methods can be used to estimate the stock prices.