一类最小二乘的自动调参问题的求解算法

An Algorithm for Solving Least Squares AutomaticParameter Tuning Problems

运用APGnc+算法思想,求解最小二乘自动调参问题.以MNIST数据集为基础,考察最小二乘自动调参在分类问题中的应用.此外,数值实验结果表明本文的算法比已有的方法快.

In this paper,the APGnc+algorithm is used to solve the least squares automatic parameter tuning problem.Based on the MNIST data set,we consider the application of least squares automatic parameter tuning in classification problems.In addition,the numerical experimental results show that the algorithm in this paper is faster than the existing methods.