Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 306 - SPEED: SPAAC SESSION II
Type: Topic-Contributed
Date/Time: Wednesday, August 11, 2021 : 3:30 PM to 5:20 PM
Sponsor: Section on Statistical Learning and Data Science
Abstract #318547
Title: A Comparative Study of Time Series Forecasting Using Deep Learning Methods
Author(s): Simachew Endale Ashebir* and Seong-Tae Kim
Companies: North Carolina A&T State University and North Carolina A&T State University
Keywords: Time Series; Forecasting; ARIMA- GARCH; Deep learning ; Artificial neural network

Time series forecasting plays a crucial role in many areas such as economics, finance, business, engineering, and natural sciences. Because of the complex behavior of stochastic dynamic time series, accurate forecasting is a big challenge. In this study, we incorporate machine learning and deep learning approaches to explore their potential beyond the traditional ARIMA framework of forecasting. We present a comparison study of ARIMA, machine learning, and deep learning methods for a real example. In this work, we implement ARIMA, ARIMA-GARCH, Random Forest, Support Vector Machine, Artificial neural network, Deep neural network, Recurrent neural network, Long short-term memory algorithms. We apply the selected algorithms to the daily stock market data. The study will demonstrate the forecasting performance of the selected algorithms and discuss how hyperparameter choices affect the performance. Our findings will provide flexible forecasting options in addition to widely used approaches.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2021 program