Water level anomaly detection and early warning method based on machine learning

The invention discloses a water level anomaly detection and early warning method based on machine learning, and belongs to the technical field of water level early warning. The method comprises the steps that firstly, weather information, historical logs, a land hydrological map and image data and environment information around a river in a designated area are collected, the river is divided into river channels in the land hydrological map, a training set is extracted from the historical logs through a machine learning algorithm, and relation parameters and risk coefficients of the river channels are obtained through analysis; secondly, an abnormal river channel and a marked river channel are divided according to the risk coefficient, image data and environment information around the marked river channel are analyzed, a relation formula is established, the water level change amplitude of the abnormal river channel in a period of time in the future is predicted in combination with meteorological information, a