<p>Roof falls are the main calamities that happen in underground coal mines. These falls seriously impression workers in terms of injuries, fatality, and production disruptions. Hence, a machine learning (ML)-based strategy is important to forecast roof falls in underground coal mines. The ML-based algorithms are very useful for foreseeing the roof convergence in the underground mine workings to advance the safety of operations. A suitable ML-based algorithm named Regression has been developed for underground mines to alleviate problems associated with roof falls. This paper mainly focuses on the development of an ML-based algorithm that has been deployed to monitor roof convergence in underground mines. Many parameters influence the roof convergence in underground mines. However, depth, retreat distance, and RMR parameters significantly impact roof convergence. In the experiment, we employed a machine learning paradigm called multiple Regression to measure the impact of the abovementioned parameters on the roof convergence. We measure the impact by considering the different combinations of the parameters. Then, the impact of the parameters is estimated by considering all the parameters simultaneously. The drafted results show that the depth, retreat distance, and RMR parameters greatly influence the predicted value of roof convergence.</p>