合成少数类过采样过滤器方法在二手车推荐中的应用

 由于二手车推荐的数据集具有非平衡特性，因此，二手车推荐可视为非平衡分类问题，可借助解决非平衡分类问题的方法来实现二手车推荐。本文对非平衡数据分类的数据集重构进行研究，通过分析合成少数类过采样方法(Synthetic Minority Over-sampling Technique, SMOTE)的特点与不足，提出合成少数类过采样过滤器方法(Synthetic Minority Over-sampling Technique Filter, SmoteFilter)，对SMOTE方法合成样本进行过滤，减少合成样本中的噪声数据，提高训练样本“质量”。使用支持向量机对SMOTE合成的数据和SmoteFilter合成的数据进行实验对比，结果表明SmoteFilter方法相较传统的SMOTE过采样方法，提高了二手车推荐中少数类的预测精度，提升了对二手车推荐的整体预测性能。

Used-car Recommendation Based on Synthetic Minority Over-sampling Technique Filter

 Due to the fact the used-car data have unbalanced characteristics, recommendation of used-cars boils down to unbalanced data classification problem and it can be solved with the unbalanced classification methods. In this paper, with the focus on reconstruction of the trainning data set and by an analysis of characteristics and deficiency of the SMOTE over-sampling method, we propose the Synthetic Minority Over-sampling Technique Filter, or SmoteFilter for short. It works by filtering the data generated by SMOTE over-sampling and reduces the noise in generated data. Based on support vector machine using data generated by SMOTE and SmoteFilter, the experimental study shows that SmoteFilter method has better effect on predicting accuracy of minority class than the SMOTE method, improving the prediction performance of vehicle recommendation.