Abstract: An algorithm for target threat assessment based on Back Propagation (BP) neural network optimized by Modified Particle Swarm Optimization (MPSO) is proposed to improve the prediction accuracy of target threat. In this MPSO algorithm, mutation operator and optimization for several parameters are introduced in PSO to avoid the particle plunging into the local optimization. The MPSO algorithm is employed to optimize the initial weights and thresholds of the BP neural network. Then the BP neural network optimized by MPSO is trained by training sets of different sample sizes. 60 sets of target threat data are adopted to test the performance of MPSO-BP in target threat prediction. Experimental results show that the prediction accuracy of target threat assessment algorithm based on MPSO-BP is higher than that based on some traditional algorithms, which proves the efficiency of the proposed algorithm in solving target threat assessment problem in spite of the small sample size of training set.

Key words: information processing technology, threat assessment, particle swarm optimization(PSO) algorithm, BP neural network, parameter optimization