摘要
SVM是一种新的理论依据充分的机器学习算法.将SVM引入水的黏度测定实验,并将SVM寻求的"最优回归函数"与最小二乘法寻求的"最佳函数"进行了对照,结果发现SVM效果较好,从而说明SVM是一种值得引进的物理实验数据分析方法.
<Abstrcat>SVM is introduced into the experiment of measuring the coefficient of viscosity. Compared with the optimal function obtained by the least square method, the regression function derived by SVM has better (effects). This demonstrates that it is worthy of introducing SVM data analysis method in physics experiments.
出处
《大学物理》
北大核心
2005年第6期40-43,共4页
College Physics
基金
国家自然科学基金资助项目(60175023)
安徽省优秀青年科技基金资助项目(04042069)
关键词
机器学习
SVM
最小二乘法
黏度
machine learning
SVM
least square methods
coefficient of viscosity
