A predictive model for climate change using advanced machine learning algorithms in Egypt | ||
| Journal of Artificial Intelligence in Engineering Practice | ||
| Volume 2, Issue 2, November 2025, Pages 1-21 PDF (1.6 M) | ||
| Document Type: Original Article | ||
| DOI: 10.21608/jaiep.2025.415474.1020 | ||
| Authors | ||
| Yasmeen Ali Shalaby* ; Asmaa Saeed Embark | ||
| Business Information System, El Gazeera High Institute For Computer & Management Information System, Cairo 11518, Egypt | ||
| Abstract | ||
| This study evaluates advanced machine learning (ML) models for forecasting daily average temperatures in Egypt, using a dataset from one of the world’s most climate databases, the GHCN-D of the NCEI under NOAA (United States). The dataset spans nine years (January 1, 2015 - December 31, 2023) and consists of 73,562 daily records from 23 climate stations across Egypt, covering eight climate features. A comparative analysis was conducted between LSTM networks and other algorithms, involving XGBoost, Random Forest, Gradient Boosting, Support Vector Regression, ARIMA, and Linear Regression. The LSTM model represented clear superiority achieving R² = 0.97, MAE = 0.07 °C, and MSE = 0.01 °C², strongly outperforming all other models, specifically in capturing long-term temporal dependencies in time series of climate. The study forecasted Egypt’s daily average temperature for August 2025. The results present a steady upward trend from about 33°C on August 1 to around 34.4°C on August 29. This gradual height aligns with peak summer in Egypt’s desert climate, which may indicate seasonal broader climate variance influences, consistent with the previous predictions of increasing temperatures. The results confirm that deep learning, specifically LSTM, presents improvements over traditional methods for temperature forecasting, providing a highly accurate predictive model. Our research contributes to advancing climate change capabilities in Mediterranean and dry regions, with practical influence for agricultural planning, environmental monitoring, and climate adaptation. | ||
| Keywords | ||
| Prediction; machine learning; regression; climate change; Egypt | ||
|
Statistics Article View: 12 PDF Download: 12 |
||
