Prediction of Apple Stock Price - Comparing the Linear and Nonlinear Regression

Abstract

Predicting future movements of a stock could be usually better performed within multiple models of analysis. However, different methods of prediction will show completely varied results even in case of a same stock. Due to varied logics and progression process, each model will return different results. These results could have implied certain aspects or properties that the stock owns, and they would reflect keys to acknowledge analyzers and decision makers to in practical. It is imperative for analyzers or decision makers to compare results from different models and algorithms. This paper performs a comparative statement of how would different prediction models or algorithms differentiate in results of predicting a stock price. Furthermore, the researcher will include a comparation of applying separately linear regression and Nonlinear Regression, to analysis on the stock Apple (NASDAQ: AAPL) and compare their prediction outcomes. In addition, the researcher will evaluate the performance according to RMSE (Root-Mean-Square-Deviation) results.