Please use this identifier to cite or link to this item:
Full metadata record
DC FieldValueLanguage
dc.contributor.authorPichayakone Rakphoen_US
dc.contributor.authorWoraphon Yamakaen_US
dc.contributor.authorRungrapee Phadkanthaen_US
dc.description.abstractThis paper aims at analyzing the energy price volatility forecasts for crude oil, ethanol, and natural gas. Several hybrid Artificial Neural Networks (ANN)-GARCH models consisting of ANN-GARCH, ANN-EGARCH, and ANN-GJR-GARCH models are introduced. However, a challenge in the ANN design is the selection of activation function. Thus, various forms of activation function, namely logistic, Gompertz, tanh, ReLU and leakyReLU are also considered to analyze the increase in the hybrid models’ predictive power. In our investigation, both in-sample and out-of-sample analysis are used and the results provide the strong evidence of the higher performance of the hybrid-ANN-GARCH compared to the single GARCH-type models. However, when five activation functions are applied over the parameters, the results tend to be similar, indicating the robustness of our forecasting results.en_US
dc.subjectComputer Scienceen_US
dc.titlePredicting Energy Price Volatility Using Hybrid Artificial Neural Networks with GARCH-Type Modelsen_US
dc.typeBook Seriesen_US
article.title.sourcetitleLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)en_US
article.volume13199 LNAIen_US Mai Universityen_US
Appears in Collections:CMUL: Journal Articles

Files in This Item:
There are no files associated with this item.

Items in CMUIR are protected by copyright, with all rights reserved, unless otherwise indicated.