I'm new in R. I have an original dataset with 25771 variables and 118 samples. I already performed feature selection and split the dataset into 70 30 so i have 82 samples in my training data and 36 in my testing data. Then, I performed SMOTE resampling on my training data only and get a new dataset with 25771 features and 144 samples. By using the output of top features from feature selection result with varying number of features such as 10,20,30 until 100 have been undergoing the classifier package to perform SVM. I already performed SVM using linear and polynomial kernel and then compute the accuracy. I got 91.57% for SVM Linear using 100 features and 75% for SVM with polynomial kernel. But why it doesn't work with RBF kernel? I only get 20% of accuracy using RBF kernel. What should I do next to improve the accuracy of SVM using RBF kernel?
Dear Andi Khalil Gibran Basir ,
Look the link, maybe useful.
http://www.rroij.com/open-access/improving-the-classification-accuracy-using-support-vector-machines-svms-with-new-kernel-1-7.php?aid=37990
Regards,
Shafagat
Why do you expect to have a good result with RBF kernel? Just because you had a good accuracy with linera SVM?
In my experiments, the regularization parameter C often affects the RBF kernel's accuracy. Maybe you can try varying C e.g. from -10^5 to +10^5.
Just try varying "gamma" and "c" parameters. In according to data distribution first set these parameters near the upper and lower bound and then tune them to reach best configuration. Sometimes gamma should be close to 0 like 0.0001 and C Vero close to 1 like 0.9.
Keep in mind that higher gamma and c values and could cause overfitting.
If your Data Set is imbalanced you should change default value of parameter "weight". Sometimes it is necessary to change bias value.
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Physiological Processes
- Stress