摘要
随着我国现代技术的不断发展,手写数字识别技术中需要处理的数据样本量及特征指标大幅增加,也对处理数据模型和方法提出了更高的要求。支持向量机(SVM)、逻辑回归模型(LR)及决策树模型(DT)等新兴机器的学习方法虽然能够处理小样本等分类问题,但在处理多样本多特征数据时分类精度还有待改进。因此,本文以提升模型处理多样本和多特征指标数据集的分类预测性能为目标,采用Stacking方法对不同的基础模型进行集成应用分析。
With the booming of digital technology in China,the number of data samples and feature indicators that need processing has increased significantly,which has put forward higher requirements for data processing models and methods.Emerging machine learning models such as support vector machine(SVM),logical regression model(LR)and decision tree model(DT)are more suitable for handling the classification of small samples,however,the classification accuracy is not enough when dealing with multi-sample and multifeature data.To solve this issue,this paper aims to employ Stacking approach to verify the integrated application of different models with the goal of improving the classification prediction performance of models dealing with multi-sample and multi-feature index datasets.
作者
符新伟
王舒可
Fu Xinwei;Wang Shuke(International Business School of Yunnan University of Finance and Economics,Kunming 650000)
出处
《中阿科技论坛(中英文)》
2022年第11期124-128,共5页
China-Arab States Science and Technology Forum