Statistics and Machine Learning: Regression and Forecasting

Under the direction of Dr. Julie Clark Statistics and machine learning are two methods for analyzing and modeling data. A dependent variable\u27s relationship to one or more independent variables is described through regression. The forecasting technique is used for predicting future results based on historical data. This study compares two approaches—regression and forecasting—along with statistics and machine learning. Regression and forecasting methods have been explored in numerous studies, but the results vary depending on the problem, the data, and the technology, so it is important to continue investigating regression and forecasting algorithms. We study the statistical model of ordinary least squares (OLS) and the machine learning model of gradient boosting regression (GBR) for the linear regression technique. For forecasting utilizing the Markov chain, we investigate the statistical model of the Autoregressive Integrated Moving Average (ARIMA) and the machine learning model of the Hidden Markov Model (HHM). Although regression and forecasting are applications of both machine learning and statistical models, there are some distinctions in their algorithms and data approaches. The choice between statistics and machine learning will depend on the specific problem, data, and available resources