针对小规模数据集的多模型融合算法研究

目前,对小规模数据集进行预测时,主要使用传统机器学习算法,但传统单一模型预测效果不能达到预期准确率,且无法兼顾多项评价指标。因此,文中以小规模数据集为研究对象,融合决策树、逻辑回归、支持向量机三类模型,提出了一种多模型融合算法,并分析了其在小规模数据集上的应用效果。首先,简述了决策树、逻辑回归和支持向量机的算法原理;其次,使用决策树、逻辑回归和支持向量机作为基学习器并完成单独训练,将各模型输出结果用于下一阶段模型输入,同时使用最大似然估计迭代优化参数,从而完成多模型融合过程;最后,对数据集进行分析和处理,通过实验与单一模型进行指标对比。实验结果表明,多模型融合算法在预测精确率、召回率、准确率等方面有明显提升。

Research on Multi-model Fusion Algorithm for Small Scale Data Sets

At present,traditional machine learning algorithms are mainly used in the prediction of small-scale data sets,but the traditional single model cannot reach the expected accuracy in prediction effect and cannot take into account multiple evaluation indexes.Therefore,taking the small-scale data sets as research objects and integrating decision tree,logistic regression and support vector machine,we propose a multi-model fusion algorithm and analyze its application effect on small-scale data sets.Firstly,the algorithm principle of decision tree,logistic regression and support vector machine is briefly described.Secondly,decision tree,logistic regression and support vector machine are used as the base learner and the individual training is completed.The output results of each model are used for the model input in the next stage,and the maximum likelihood estimation is used for iterative optimization parameters to complete the multi-model fusion process.Finally,the data sets are analyzed and processed,and the indicators are compared with the single model through experiments which show that this algorithm has a significant improvement in prediction precision,recall rate and accuracy.