通过建立线性积分模型,研究犯罪嫌疑人的分类分级问题,实现对犯罪行为的预警。将积分模型参数的确定归结为目标函数非解析的最佳参数识别问题,并利用遗传算法得到最佳参数。为了克服传统遗传算法的早熟现象和局部收敛问题,文中提出了一种具有改进的交叉算子和变异算子的遗传算法。改进的遗传算法可加快算法收敛速度,从而缩短寻找最优解的时间,提高算法的效率。数值实验结果表明,改进遗传算法用于犯罪嫌疑人的分类分级问题寻找最佳参数有更高的运行效率,建立的积分模型对犯罪嫌疑人的分类分级具有较好的准确率。%By establishing the linear cumulative model,it studies the classification of the criminal suspects to realize the early warning of criminal activity. The cumulative model parameters determined come down to solving a target function non-analytic of the parameters i-dentification problem,and use genetic algorithm to solve this model getting the optimal parameters. In order to overcome premature and local convergence of traditional genetic algorithm,it puts forward an improved genetic algorithm with a modified crossover operator and mutation operator in this paper. Improved genetic algorithm can speed up the convergence rate,thus shortening the time to find the optimal solutions and improving the efficiency of full instructions. The experimental result shows that improved genetic algorithm applied to crimi-nal suspects has better operating efficiency to solve optimal parameters classification problem,and this cumulative model has good accura-cy.
展开▼