基于机器学习的轿车级别评定

Car level evaluation based on the machine learning

建立有效的数学模型可以提高消费者判定轿车级别的准确率。首先,将数据集分为训练数据集和测试数据集,针对训练数据集分别采用支持向量机算法、随机森林算法、k-近邻算法以及朴素贝叶斯算法建立4种模型。其次,用测试数据集比较和分析模型的性能,详细分析了每个模型的优缺点,以准确率作为评价指标,并根据消费者的购买需求,选择相应的模型。

The accuracy for level of car is improved through the establishment of effective mathematical model.First,the da-ta is divided into training set and testing set.Moreover,based on support vector machine,random forests,k-nearest neigh-bor and naive Bayesian algorithms,four models are established in view of the training set,respectively.Next,comparing the performance of the models and analyzing the advantages and disadvantages of each model,the appropriate model is choose according to the accuracy rate and the consumer′s purchase requirement.