You are on page 1of 2

Abstract The purpose of the work is about machine learing, creating machine models using support vector machine

algorithm and examine the test errors and training errors for the given dataset.

Introduction

A support vector machine constructs a hyperplane or set of hyperplanes in a high- or innite-dimensional space, which can be used for classication, regression, or other tasks. Classication is the problem of identifying to which of a set of categories (sub-populations) a new observation belongs, on the basis of a training set of data containing observations (or instances) whose category membership is known.

Methods

Machine model is created using SVM algorithm. Training and Testing les are created from the input datale. Training and Test data sets are created randomly at various percentages using the following algorithm. Assuming that p is the percentage of training le and the remainder in the test le, 1. Read a line from the le 2. Generate a uniform random number from (0,1), k 3. If k>p write the line to testle else write it to train le 4. Creating the training and testing dataset from the input data le By using the training dataset machine model is created, from the model the training and testing errors are calculated. The algorthim is as follows 1. Creating the training and testing dataset from the input data le 2. Creating the model using command svm-train <traing data set> <model name> 3. Creating the accuracy of the model using the test dataset using the command svm-predict <test dataset> <model name> <results le> 4. Creating the accuracy of the model using the training dataset using the command svm-predict <training dataset> <model name> <results le>

5. Calculating the test error and traing error from the accuracy obtained from the above results

You might also like