摘要
基于 SVM分类器在模式识别问题中有独特的优势 ,本文通过对标准 SVM模型的改造 ,提出了一种新的简单的数据分类方法 .理论分析和实验表明 ,该方法与标准 SVM分类方法相比具有处理大规模数据识别的能力且保持较高的样本识别率 ,节省存储空间等优势 .
Based on SVM Classifier has great advantage in the pattern recognition. A new extremely simple procedure is presented for data classification by changing standard SVM in this paper. Comparing with standard SVM, the algorithm can easily handle large datasets, guarantee high test set correctness and save storage space by analyses and experiments.
出处
《数学的实践与认识》
CSCD
北大核心
2003年第12期61-65,共5页
Mathematics in Practice and Theory
