SVM learning

Using SVM for Real-Life Problems

%matplotlib inline

import pandas as pd
import numpy as np
from sklearn import svm
import matplotlib.pyplot as plt
import seaborn as sns; sns.set(font_scale=1.2)

data = pd.read_csv('house_sizes_prices_svm.csv')

sns.lmplot('size', 'price',
           data=data,
           hue='sold',
           palette='Set2',
           fit_reg=False,
           scatter_kws={"s": 50});

X = data[['size','price']].values
y = np.where(data['sold']=='y', 1, 0) #--1 for Y and 0 for N---
model = svm.SVC(kernel='linear').fit(X, y)

sns.lmplot('size', 'price',
           data=data,
           hue='sold',
           palette='Set2',
           fit_reg=False,
           scatter_kws={"s": 50});

#---min and max for the first feature---
x_min, x_max = X[:, 0].min() - 1, X[:, 0].max() + 1

#---min and max for the second feature---
y_min, y_max = X[:, 1].min() - 1, X[:, 1].max() + 1

#---step size in the mesh---
h = (x_max / x_min) / 20

#---make predictions for each of the points in xx,yy---
xx, yy = np.meshgrid(np.arange(x_min, x_max, h),
                     np.arange(y_min, y_max, h))

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

#---draw the result using a color plot---
Z = Z.reshape(xx.shape)
plt.contourf(xx, yy, Z, cmap=plt.cm.Blues, alpha=0.3)

plt.xlabel('Size of house')
plt.ylabel('Asking price (1000s)')
plt.title("Size of Houses and Their Asking Prices")

Text(0.5,1,'Size of Houses and Their Asking Prices')

def will_it_sell(size, price):
    if(model.predict([[size, price]]))==0:
        print('Will not sell!')
    else:
        print('Will sell!')

#---do some prediction---
will_it_sell(2500, 400)  # Will not sell!
will_it_sell(2500, 200)  # Will sell!

Will not sell!
Will sell!