【摘要】 快速且高效地识别钓鱼网站是非常有必要的，可以帮助人们有效避免钓鱼网站引起的网络安全威胁。提出一种基于Stacking集成模型对钓鱼网站进行识别的模型。首先对数据预处理，再利用XGBoost算法进行最优特征集筛选，建立单一模型和Stacking集成模型，同时使用分层交叉验证和网格搜索对算法参数进行调节。实验结果表明，基于Stacking集成结构对钓鱼网站的识别准确率达到了97.96%,AUC值为0.9801，该方法相比其他单一分类器具有更高的识别能力。更多还原

【Abstract】 It is very necessary to identify phishing websites quickly and efficiently, which can help people effectively avoid network security threats caused by phishing websites. This paper presents a Stacking integrated model for phishing site identification. The data was preprocessed, the XGBoost algorithm was used for optimal selection and Stacking, a single model and Stacking integration model were established, and the algorithm parameters were adjusted using layered cross-validation and grid search. The experimental results show that the Stacking integration has a higher recognition accuracy of 97.96%, with an AUC value of 0.9801,compared with other single classifiers.更多还原