Learning
6 models are used in the learning steps: K-Nearest Neighbor, perception, decision tree, support vector machine, multi-layer perception and Na?ve Bayesian. I designed a RBF classifier and MLP classifier at first but they are too slow for the reason that matrix manipulation hasn’t been designed carefully, so I use the function in the library instead. Parameter determination for these classifiers are:
①K-NN
When k≥5，the accuracy trends to be stable, so k=5
②Decision tree
Maxcrit is used as binary splitting criterion.
③MLP
5 units for hidden is enough
最后
以上就是虚拟短靴最近收集整理的关于华南理工 计算机大作业论文,华南理工大学《模式识别》大作业报告的全部内容，更多相关华南理工内容请搜索靠谱客的其他文章。
本图文内容来源于网友提供，作为学习参考使用，或来自网络收集整理，版权属于原作者所有。
![[DEBUG] matlab，matconvnet中常遇到的报错](/uploads/reation/bcimg4.png)
发表评论 取消回复