A method for determining parameter weight early warning model based on reinforcement learning

Abstract

To solve the problem of low efficiency of early warning in public security emergencies, this paper proposes a method for determining parameter weight in public events early warning model which was based on reinforcement learning. Firstly, using the calibrated conflict early warning label, using reinforcement learning algorithm to build the public early warning event model; secondly, through iterative training to obtain the arrival path of the agent to the abnormal sequence, that is, the public early warning event; finally, by analyzing the weight parameters in the neural network, to determine the early warning event. Simulation showed that under this algorithm, convergence happened when the number of steps was in the range from 500 to 800, 37.5% smaller than that when using the original data. This result of the experiment demonstrated that this method greatly improved the efficiency of early warning for public incidents.