Ensuring unbiasedness: foundational insights into integrating GSTARIMA and DNN models for rainfall prediction.
Saved in:
| Title: | Ensuring unbiasedness: foundational insights into integrating GSTARIMA and DNN models for rainfall prediction. |
|---|---|
| Authors: | Munandar, Devi (AUTHOR), Ruchjana, Budi Nurani (AUTHOR), Abdullah, Atje Setiawan (AUTHOR), Pardede, Hilman Ferdinandus (AUTHOR) |
| Source: | Connection Science. Dec 2025, Vol. 37 Issue 1, p1-21. 21p. |
| Subjects: | Maximum likelihood statistics, Time series analysis, Artificial neural networks, Precipitation forecasting, Objectivity, Box-Jenkins forecasting, Spatial data structures, Machine learning |
| Abstract: | The GSTARIMA (Generalied Space–Time Autoregressive Integrated Moving Average) model is commonly used to analyse time series and spatial data with temporal and spatial dependencies. This paper focuses on estimating the autoregressive and moving average parameters of the GSTARIMA model using Maximum Likelihood Estimation (MLE). We theoretically demonstrate the unbiasedness of these estimates, proving that the expected values of the estimates match the true parameters. Empirical experiments further verify this property, both before and after applying Deep Neural Network (DNN) interventions to correct model errors. The results show that the parameter estimates remain unbiased, and error properties (zero mean and constant variance) are preserved even after DNN processing. This study highlights the robustness of MLE in providing unbiased estimates within the GSTARIMA framework, even when integrated with machine learning techniques. [ABSTRACT FROM AUTHOR] |
| Copyright of Connection Science is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Psychology and Behavioral Sciences Collection |
|
Full text is not displayed to guests.
Login for full access.
|
|
| Abstract: | The GSTARIMA (Generalied Space–Time Autoregressive Integrated Moving Average) model is commonly used to analyse time series and spatial data with temporal and spatial dependencies. This paper focuses on estimating the autoregressive and moving average parameters of the GSTARIMA model using Maximum Likelihood Estimation (MLE). We theoretically demonstrate the unbiasedness of these estimates, proving that the expected values of the estimates match the true parameters. Empirical experiments further verify this property, both before and after applying Deep Neural Network (DNN) interventions to correct model errors. The results show that the parameter estimates remain unbiased, and error properties (zero mean and constant variance) are preserved even after DNN processing. This study highlights the robustness of MLE in providing unbiased estimates within the GSTARIMA framework, even when integrated with machine learning techniques. [ABSTRACT FROM AUTHOR] |
|---|---|
| ISSN: | 09540091 |
| DOI: | 10.1080/09540091.2025.2507179 |
