딥러닝기법과 최소자승법을 활용한 미국 주식시장 예측 연구
Daily Stock Prices Forecast in the US Stock Market with Deep Learning Method and Least Square Method
이익선(동아대학교)
37권 2호, 19~31쪽
초록
It has been widely acknowledged that stock prices are influenced by various kinds of information and it is very difficult to measure the impact of such information. This paper selected eight industries in the US stock market, and a total of 249 companies within eight industries, and developed a methodology to predict the stock prices of those companies. From eight industries such as Basic Materials, Conglomerates, Consumer Goods, Finance, Healthcare, Industrial Goods, Services, Utilities, 3,523 daily stock price data were collected from 2002 to 2015. To predict the stock price of 249 companies, 21 major indices were used as independent variables. This paper used the deep learning and the least square method to predict the fluctuations in the US stock market.
Abstract
It has been widely acknowledged that stock prices are influenced by various kinds of information and it is very difficult to measure the impact of such information. This paper selected eight industries in the US stock market, and a total of 249 companies within eight industries, and developed a methodology to predict the stock prices of those companies. From eight industries such as Basic Materials, Conglomerates, Consumer Goods, Finance, Healthcare, Industrial Goods, Services, Utilities, 3,523 daily stock price data were collected from 2002 to 2015. To predict the stock price of 249 companies, 21 major indices were used as independent variables. This paper used the deep learning and the least square method to predict the fluctuations in the US stock market.
- 발행기관:
- 한국경영과학회
- 분류:
- 경영학