论文部分内容阅读
提出一种基于遗传算法的顾客行为特征提取算法.首先,采用Tan imoto相似度来度量顾客间购买行为,并设计遗传聚类算法对顾客群体进行划分,把具有相似购买行为顾客聚集为一类.然后,针对不同顾客群体的购买行为特征,设计一种基于遗传算法的多种群特征提取方法,从各个子群体中发现顾客的购买行为的知识.为了增强种群内部协同进化能力和规则质量,我们采用最近邻替代遗传策略和局部搜索策略.使用实际零售数据集对整个算法进行验证,并与经典的Apriori算法进行比较.实验结果表明该算法在不需要产生频繁项集的情况下,可较高效生成精简规则集,在规则形式方面也更加灵活.最后,对实验结果进行详细分析.
This paper proposes a genetic algorithm based customer behavior feature extraction algorithm.First, Tanimoto similarity is used to measure interpersonal purchase behavior, and genetic clustering algorithm is designed to segment customer groups, customers with similar purchasing behavior are aggregated into one category. Then, according to the characteristics of different customers’ purchasing behavior, a multi-population feature extraction method based on genetic algorithm is designed to find out the customer’s purchasing behavior knowledge from each sub-population.In order to enhance the co-evolvement ability and rule quality within the population, we adopt Nearest neighbor alternative genetic strategy and local search strategy.The whole algorithm is verified by the actual retail data set and compared with the classical Apriori algorithm.The experimental results show that the proposed algorithm can generate more efficiently without generating frequent itemsets Streamlining rule sets is also more flexible in the form of rules.Finally, the experimental results are analyzed in detail.