is it possible to get 100% classification accuracy in SVM classifier lerner app in matlab?

5 次查看（过去 30 天）
kc
kc 2020-2-1
回答： kc 2020-2-3
I am getting 100% accuracy in almost every model in classifier leraner app MATLAB . Initially i gt few NaN entries in ma feature set ,but after removing them i m still getting it same almost 5000 samples and 300 features.
Plz suggest some debug methods.
  3 个评论
显示 1更早的评论
kc
kc 2020-2-2
Sir i have divident the data randomely in 70-30 . Given 70 with labels in Classifier app, and then extract the model to test the remaining 30 , which is not given to classifier before , and still its giving high accuracy(almost same a lil lower). But i know on that dataset its almost impossible with these features. So plz guide what else i need to check?
one more important question is is it fine to use this app or i should go for code. But eventually that is same as generated code from app ? right?
plz help
Image Analyst
Image Analyst 2020-2-2
Please post a scatterplot of your data so we can see if the data points are well separated or not. Attach your data if you need more help.

请先登录，再进行评论。

请先登录，再回答此问题。

回答（2 个）

kc
kc 2020-2-3
Capture_matlabcentral.JPG
kc
kc 2020-2-3
Sir i have attched the scatter plot but data file is large so unable to do so.

类别

在 Help Center 和 File Exchange 中查找有关 Get Started with Statistics and Machine Learning Toolbox 的更多信息

标签

产品

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by