Statistical Machine Learning Model and Commodity Futures Volatility Information for Financial Stock Market Forecasting

Denis A. Pustokhin¹, Irina V. Pustokhina*^,2

¹ Department of Logistics and Marketing, Faculty of Economics and Business, Financial University under the Government of the Russian Federation, Leningradskiy Prospekt 55, Moscow 125993, Russian

² Department of Logistics, State University of Management, Moscow 109542, Russian

Emails: dapustokhin@fa.ru; iv_pustokhina@guu.ru

Abstract

A country's economy and social structure are greatly influenced by the stock market. It is extremely difficult for investors, expert analysts, and scholars in the financial industry to forecast the stock market accurately because of the pretty unstable, parametric, non-linear dynamical, and unstable character of stock price time series. In the financial sector, stock market forecasting is a critical activity and a prominent study subject because stock market investments carry greater risk. It's conceivable, however, to reduce most of the risk through the development of computationally intelligent approaches. This paper introduces the support vector machine regression to make a model forecasting the stock market financial.

Keywords: SVM; Machine Learning; Forecasting; Regression