2) Decision Tree.txt

Decision Tree 

import numpy as np 
import pandas as pd 
import matplotlib.pyplot as plt 
data = pd.read_csv("ads.csv") 
data.head() 
data.isnull().any() 
x = data.iloc[:,1:4].values 
y = data.iloc[:,4:5].values 
x 
y[:10] 
from sklearn.preprocessing import LabelEncoder 
lb = LabelEncoder() 
x[:,0] = lb.fit_transform(x[:,0]) 
x 
from sklearn.model_selection import train_test_split as tts 
x_train,x_test,y_train,y_test = tts(x, y, test_size = 0.1,random_state=0) 
x_train 
x_test[:10] 
from sklearn.tree import DecisionTreeClassifier 
dt = DecisionTreeClassifier(criterion='entropy') 
dt.fit(x_train,y_train) 
y_pred = dt.predict(x_test) 
y_pred 
from sklearn.metrics import accuracy_score 
accuracy_score(y_test,y_pred) 
from sklearn.metrics import confusion_matrix 
confusion_matrix(y_test,y_pred) 
import sklearn.metrics as metrics 
fpr, tpr ,threshold = metrics.roc_curve(y_test,y_pred) 
roc_auc = metrics.auc(fpr,tpr) 
roc_auc 
plt.plot(fpr,tpr) 
plt.xlim([0,1]) 
plt.ylim([0,1]) 
plt.style.use("dark_background") 
from sklearn.tree import export_graphviz 
export_graphviz(dt, out_file ='tree.dot', 
feature_names = ["Gender","Age","Salary"], class_names = ['0','1'], 
rounded = True, proportion = False, precision = 2, filled = True) 
!pip install pydotplus 
!conda install graphviz 
!dot tree.dot -Tpng -o image.png