Minimum Message Length in Hybrid ARMA and LSTM Model Forecasting
Modeling and analysis of time series are important in applications including economics, engineering, environmental science and social science. Selecting the best time series model with accurate parameters in forecasting is a challenging objective for scientists and academic researchers. Hybrid model...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-11-01
|
| Series: | Entropy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1099-4300/23/12/1601 |
| _version_ | 1827672640928612352 |
|---|---|
| author | Zheng Fang David L. Dowe Shelton Peiris Dedi Rosadi |
| author_facet | Zheng Fang David L. Dowe Shelton Peiris Dedi Rosadi |
| author_sort | Zheng Fang |
| collection | DOAJ |
| description | Modeling and analysis of time series are important in applications including economics, engineering, environmental science and social science. Selecting the best time series model with accurate parameters in forecasting is a challenging objective for scientists and academic researchers. Hybrid models combining neural networks and traditional Autoregressive Moving Average (ARMA) models are being used to improve the accuracy of modeling and forecasting time series. Most of the existing time series models are selected by information-theoretic approaches, such as AIC, BIC, and HQ. This paper revisits a model selection technique based on Minimum Message Length (MML) and investigates its use in hybrid time series analysis. MML is a Bayesian information-theoretic approach and has been used in selecting the best ARMA model. We utilize the long short-term memory (LSTM) approach to construct a hybrid ARMA-LSTM model and show that MML performs better than AIC, BIC, and HQ in selecting the model—both in the traditional ARMA models (without LSTM) and with hybrid ARMA-LSTM models. These results held on simulated data and both real-world datasets that we considered.We also develop a simple MML ARIMA model. |
| first_indexed | 2024-03-10T04:11:08Z |
| format | Article |
| id | doaj.art-680ae6d9cb2f4eed974d6c4be31fb76a |
| institution | Directory Open Access Journal |
| issn | 1099-4300 |
| language | English |
| last_indexed | 2024-03-10T04:11:08Z |
| publishDate | 2021-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Entropy |
| spelling | doaj.art-680ae6d9cb2f4eed974d6c4be31fb76a2023-11-23T08:10:28ZengMDPI AGEntropy1099-43002021-11-012312160110.3390/e23121601Minimum Message Length in Hybrid ARMA and LSTM Model ForecastingZheng Fang0David L. Dowe1Shelton Peiris2Dedi Rosadi3Department of Data Science and Artificial Intelligence, Monash University, Clayton, VIC 3800, AustraliaDepartment of Data Science and Artificial Intelligence, Monash University, Clayton, VIC 3800, AustraliaSchool of Mathematics and Statistics, University of Sydney, Camperdown, NSW 2006, AustraliaDepartment of Statistics, Gadjah Mada University, Sleman, Yogyakarta 55500, IndonesiaModeling and analysis of time series are important in applications including economics, engineering, environmental science and social science. Selecting the best time series model with accurate parameters in forecasting is a challenging objective for scientists and academic researchers. Hybrid models combining neural networks and traditional Autoregressive Moving Average (ARMA) models are being used to improve the accuracy of modeling and forecasting time series. Most of the existing time series models are selected by information-theoretic approaches, such as AIC, BIC, and HQ. This paper revisits a model selection technique based on Minimum Message Length (MML) and investigates its use in hybrid time series analysis. MML is a Bayesian information-theoretic approach and has been used in selecting the best ARMA model. We utilize the long short-term memory (LSTM) approach to construct a hybrid ARMA-LSTM model and show that MML performs better than AIC, BIC, and HQ in selecting the model—both in the traditional ARMA models (without LSTM) and with hybrid ARMA-LSTM models. These results held on simulated data and both real-world datasets that we considered.We also develop a simple MML ARIMA model.https://www.mdpi.com/1099-4300/23/12/1601long short-term memoryminimum message lengthtime seriesneural networkdeep learningBayesian statistics |
| spellingShingle | Zheng Fang David L. Dowe Shelton Peiris Dedi Rosadi Minimum Message Length in Hybrid ARMA and LSTM Model Forecasting Entropy long short-term memory minimum message length time series neural network deep learning Bayesian statistics |
| title | Minimum Message Length in Hybrid ARMA and LSTM Model Forecasting |
| title_full | Minimum Message Length in Hybrid ARMA and LSTM Model Forecasting |
| title_fullStr | Minimum Message Length in Hybrid ARMA and LSTM Model Forecasting |
| title_full_unstemmed | Minimum Message Length in Hybrid ARMA and LSTM Model Forecasting |
| title_short | Minimum Message Length in Hybrid ARMA and LSTM Model Forecasting |
| title_sort | minimum message length in hybrid arma and lstm model forecasting |
| topic | long short-term memory minimum message length time series neural network deep learning Bayesian statistics |
| url | https://www.mdpi.com/1099-4300/23/12/1601 |
| work_keys_str_mv | AT zhengfang minimummessagelengthinhybridarmaandlstmmodelforecasting AT davidldowe minimummessagelengthinhybridarmaandlstmmodelforecasting AT sheltonpeiris minimummessagelengthinhybridarmaandlstmmodelforecasting AT dedirosadi minimummessagelengthinhybridarmaandlstmmodelforecasting |