|
Title
|
PERFORMANCE ANALYSIS OF DEEP LEARNING IN COMPARISON TO MACHINE LEARNING MODELS FOR MODELLING STOCK RETURNS TIME SERIES (EVIDENCE FROM TEHRAN STOCK EXCHANGE)
|
|
Type
|
Refereeing
|
|
Keywords
|
Financial Investment, Changes In Stock Returns, Time Series Modelling, Deep Learning, Machine Learning.
|
|
Abstract
|
This study aims at performance analysis of Deep Learning (DL) and Machine Learning (ML) models in modelling and predicting the stock returns time series, based on the return rate of previous periods and a set of exogenous variables. The data used includes the weekly data of the stock return index of 200 companies included in the Tehran stock exchange market from 2016 to 2021. Two Long Short-Term Memory LSTM) and Deep Q-Network (DQN) models as DL processes and two Random Forest (RF) and Support Vector Machine (SVR) models as ML algorithms were selected. The results showed the superiority of DL algorithms over ML models, which can indicate the existence of strong dependence patterns in these time series, as well as relatively complex nonlinear relationships with uncertainty between the determinant variables. Meanwhile, the LSTM has the highest accuracy and the least error of prediction. The RF model showed the lowest prediction accuracy by involving the highest prediction error.
|
|
Researchers
|
Nooshin Hakamipour (Referee)
|