/svmpy

Basic soft-margin kernel SVM implementation in Python

Primary LanguagePython

SVM-Python

By Dongfeng Gu (https://www.gdf.name)

Introduction

This is a basic implementation of a soft-margin kernel SVM solver in Python using numpy and cvxopt.

Usage

Training:

1. Go to the bin/svm-train, train the dataset first by passing the dimension of the X features and the directory of the dataset

2. run the program and it will generate a model.txt which will be used in the testing part

Testing:

1. Go to the bin/svm-test, test the dataset first by passing the dimension of the X features and the directory of the dataset, it will automatically search for the model.txt file.

  1. run the program and it will generate the error rate