The spontaneous combustion of coal seams is a major concern in mines, ranking among the top five most significant incidents and contributing to secondary disasters like mine gas and coal dust explosions. As mining technology advances, the detection of spontaneous coal fires has evolved, presenting new challenges for prevention and control. The likelihood of coal's spontaneous combustion varies by seam location and depends on the coal's inherent properties along with geological and mining conditions. Accurate forecasting of the risk of coal fires is crucial for the safety of coal mines and utility industry. Deep Learning (DL), a subset of Artificial Intelligence (AI), has garnered interest for its success in various real-world applications. In this study, we propose using DL techniques to predict the likelihood of spontaneous coal heating leading to mine fires. By analysing relevant parameters that trigger coal heating, we aim to develop an early warning system for taking precautionary measures.