基于变异学习的药物构效关系模式分类法

提出一类用于预测药物活性的神经元模式分类器及其神经网络进化学习新算法－－变异基遗传法。该算法仅采用变异算子进行局部优化搜索，并利用随机初始化技术使算法在局部搜索功能提高的同时仍有可能寻找到全局最优解。该研究以有机氟农药为例，将基于变异学习的神经元模式分类方法应用于药物构效关系的识别研究，实验结果表明，该方法能有效地辨识药物构效关系模式，模式分类识别正确率优于其他神经元模式分类法。

The pattern classification neural network for identifying the structure-activity relationships of drugs based on the mutation evolutionary algorithm

A nwe pattern classification method for identifying the structure\|activity relationship of drugs with the neural network based on the mutation evolutionary algorithm is presented.The mutation evolutionary algorithm named mutation\|based genetic algorithm (MGA) only uses the mutation operator for local search.To enhance the efficiency of local search,the genes that represent the variables employ different time\|varying mutation rates in MGA. Combining random restart technique with the local search strategy,the algorithm can obtain satisfactory solution in a limited time.The proposed method was used to predicting the activity of organic fluorinated pesticides for verifying the effectiveness.The study demonstrated that the predictive ability of the neural network with MGA is better than those provided by other neural network algorithms.