Disaster has been a threat to human since it appears. Forecasting these events has become a crucial approach to mitigating property damage and casualties. The rapid advancements in Artificial Intelligence (AI), particularly in Machine Learning (ML) and Deep Learning (DL), have introduced novel predictive methodologies in the disaster forecasting domain. This paper focus on three main disasters as floods, landslides, and earthquakes, comprehensively analyze prediction methods via research nature and methodology. We present existed pre-diction methods and summarizes research achievements since 2019, with a particular emphasis on the application and prospects of ML and DL techniques in disaster prediction. The paper also discusses recent research outcomes and anticipates future development trends in this field.