+- Python Forum (https://python-forum.io)
+-- Forum: Python Coding (https://python-forum.io/forum-7.html)
+--- Forum: General Coding Help (https://python-forum.io/forum-8.html)
+--- Thread: Improving Efficiency of SVM by various available kernels (/thread-9441.html)
Improving Efficiency of SVM by various available kernels - Sachtech - Apr-09-2018
Dear friend,
I have shared my folder here: GoogleDrive
I have made 2 codes and 1 PDF and Lower Spine Dataset from machine learning is used from kaggle.
In mittal_20_3.py file i have divided the dataset in train and test and found performance of SVM algorithm.
Now my next task is to improve efficiency of SVM, to do so,
I have two ways,
1) Implementing Kernel Trick that is defining new kernel for SVM.
2) Or making combination of various available kernels and improve efficiency, for that refer attached PDF
In second .py file i tried to manually code for SVM so that kernel can be code uniquely.
I would like to optimize the kernel.
To improve efficiency there are two ways,
1) Kernel function can be modified, which is nothing but matrix multiplication.
or
2) Use various combination of available kernel functions.
Now i m having doubt in coding, how to code for such options in python.
I hope to get some help from fellow members.