Throughout history, earthquakes have caused a lot of trauma and distress due to their devastating consequences. Due to the uniqueness of the seismic structures, each seismic event is different from other; therefore, prior estimation of these events is critical. Several researchers in the past have worked on prediction of earthquake occurrences; however, there is still room for improvement in terms of accuracy. With the reign of machine learning methods and its efficiency realized through various fields of research related to predictions, many researchers have found it to be a promising tool to enhance earthquake prediction performance. In this work, authors implemented multiple machine learning and ensemble learning models to predict the earthquake using USGS dataset. Among various algorithms including random forest, gradient boosting, tree, K-nearest neighbours and stack ensemble of these algorithms, the stack model provided best performance in terms of mean squared error (0.069) and root mean squared error (0.262).