The invention provides a personnel management and control method and
system based on
machine learning. The method comprises the following steps of S1, building a personnel archive which comprises necessary personnel identities and data collection identities, S2, acquiring real-
time data acquired by a
base station provided by a police internal operator and real-
time data acquired by front-end equipment, and cleaning and filtering the data to obtain non-repetitive and non-abnormal personnel training data, S3, based on the personnel archive data obtained in the step S1 and the training data obtained in the step S2, performing model training based on an unsupervised isolated forest method to obtain an
anomaly detection model, S4, based on the training model in the step S3, carrying out real-time prediction alarm through a flow calculation engine, displaying early warning information through GIS
visualization in a layered manner, and sending the early warning information to police officersfor disposal feedback, and S5, automatically updating and correcting the model based on the police officer disposal feedback information in the step S4. The judgment result is accurate, the function of multi-dimensional
mass data of the police
system is fully played, and the working efficiency is high.