Prediction of Peatlands Forest Fires in Malaysia Using Machine Learning

Abstract

The occurrence of fires in tropical peatlands poses significant threats to their ecosystems. An Internet of Things (IoT) system was developed to measure and collect fire risk factors in the Raja Musa Forest Reserve (RMFR) in Selangor, Malaysia, to address this issue. In this paper, neural networks with different layers were employed to predict peatland forests’ Fire Weather Index (FWI). The neural network models used two sets of input parameters, consisting of four and nine fire factors. The predicted FWI values were compared with actual values obtained from the Malaysian meteorological department. The findings revealed that the five-layer neural network outperformed others in both the four-input and nine-input models. Specifically, the nine-input neural network achieved a mean square error (MSE) of 1.116 and a correlation of 0.890, surpassing the performance of the four-input neural network with the MSE of 1.537 and the correlation of 0.852. These results hold significant research and practical implications for precise peatland fire prevention, control, and the formulation of preventive measures.

Utilizing Grid Data and Deep Learning for Forest Fire Occurrences and Decision Support: A Case Study in the Ningxia Hui Autonomous Region

“…These conditions contribute to limited rainfall, water scarcity, and a delicate ecological environment, posing challenges to both the natural environment and socioeconomic development [25]. [21,22]. Ningxia is situated within the sand control belt of northern China and falls within the ecological protection belt of the Silk Road, making it strategically important for China's ecological security [23].…”

“…NNs are more suitable for processing one-dimensional data like text or time series data. CNNs typically feature a lower parameter count, simplifying training processes even with limited datasets [21,22].…”