Parameter-optimization-of-SVM

The repository contains the R script for optimizing the parameters of SVM

Dataset - https://archive.ics.uci.edu/ml/datasets/wine

Attribute information

Input variables (based on physicochemical tests):

1 - fixed acidity

2 - volatile acidity

3 - citric acid

4 - residual sugar

5 - chlorides

6 - free sulfur dioxide

7 - total sulfur dioxide

8 - density

9 - pH

10 - sulphates

11 - alcohol

Output variable (based on sensory data):

12 - quality (score between 0 and 10)

Tasks performed

• Download the dataset

• Pre-process the dataset

• Create ten samples

• Split the samples in 70 : 30 for training and testing

• Optimise SVM using randomisation for every sample and report best accuracy and best parameters

• For the best sample plot the convergence graph

Results

sample	best_accuracy	best_kernel	best_nu	best_epsilon
S1	0.649916247906198	laplacedot	0.600165726849809	0.958756402833387
S2	0.651591289782245	laplacedot	0.162177571561188	0.727429841179401
S3	0.62751677852349	laplacedot	0.939081447897479	0.813470450695604
S4	0.628140703517588	laplacedot	0.712692788802087	0.705607258714736
S5	0		0	0
S6	0.638190954773869	laplacedot	0.305205665994436	0.376962826121598
S7	0.644891122278057	laplacedot	0.619259966304526	0.90585225680843
S8	0		0	0
S9	0.629815745393635	laplacedot	0.0905253798700869	0.21171743911691
S10	0.649328859060403	laplacedot	0.332543317927048	0.137839282630011

Convergence Graph