1. Python
  2. ML | Nonlinear SVM

ML | Nonlinear SVM



In

These classes are clearly not linearly separable. Below is a nonlinear SVM contour plot that successfully classified the IRIS dataset using RBF core.

The above figure shows the classification of the three classes of the IRIS dataset.

  1. From sklearn, we imported the SVM library.
  2. We created 3 non-linear SVM`s (RBF kernel based).
  3. Each SVM was fed with 1 class kept positive and other 2 as negative. Say, SVM1 had labels corresponding to class 1 only else all were made 0. Same for SVM2 and SVM3 respectively.
  4. Plot the contour plot of each SVM.
  5. Plot the data points.

Below is a Python implementation for the same.

import numpy as np 

import pandas as pd 

import matplotlib.pyplot as plt 

from matplotlib import style

from sklearn.svm import SVC 

 

style.use ( `fivethirtyeight` )

  
# create grid

def make_meshgrid (x , y, h = . 02 ):

x_min, x_max = x. min () - 1 , x. max () + 1

y_min, y_max = y. min () - 1 , y. max () + 1

xx, yy = np.meshgrid (np.arange (x_min, x_max, h), np.arange (y_min, y_max, h))

return xx, yy

 
# build outlines

def plot_contours (ax, clf, xx, yy, * * params):

  Z = clf.predict (np.c_ [xx.ravel (), yy.ravel ()])

Z = Z.reshape (xx.shape)

out = ax.contourf (xx, yy, Z, * * params)

  return out

  

color = [ `r` , ` b` , `g` , ` k` ]

 

iris = pd.read_csv ( " iris-data. txt " ). values ​​

  

 

features = iris [ 0 : 150 , 2 : 4 ]

level1 = np.zeros ( 150 )

level2 = np.zeros ( 150 )

level3 = np.zeros ( 150 )

 
# level1 contains 1 for class1 and 0 for all others.
# level2 contains 1 for class2 and 0 for all others.
# level3 contains 1 for class3 and 0 for all others.

for i in range ( 150 ):

if i & gt; = 0 and i & lt; 50 :

  lev el1 [i] = 1

elif i & gt; = 50 and i & lt; 100 :

level2 [i] = 1

elif i & gt; = 100 and i & lt; 150 :

level3 [i] = 1

  # create 3 svm with rbf cores

svm1 = SVC (kernel = `rbf` )

svm2 = SVC (kernel = ` rbf` )

svm3 = SVC (kernel = `rbf` )

 
# fits every svm
svm1.fit (features, level1)
svm2.fit (features, level2)
svm3.fit (features, level3)

 < / p>

fig, ax = plt.subplots ()

X0, X1 = iris [:, 2 ], iris [:, 3 ]

xx, yy = make_meshgrid (X0, X1)

  
# build outlines

plot_contours (ax, svm1, xx, yy, cmap = plt.get_cmap ( ` hot` ), alpha = 0.8 )

plot_contours (ax, svm2, xx, yy, cmap = plt.get_c map ( `hot` ), alpha = 0.3 )

plot_contours (ax, svm3, xx, yy, cmap = plt.get_cmap ( `hot` ), alpha = 0.5 )

 

color = [ `r` , `b` , ` g` , `k` ]

  

for i in range ( len ( iris)):

plt.scatter (iris [i] [ 2 ], iris [i] [ 3 ], s = 30 , c = color [ int (iris [ i] [ 4 ])])

plt.show ()