import numpy as np import pandas as pd

df = pd.read_csv("C:/Users/Umaima/Untitled Folder/bank_data.csv", index_col=0, na_values=["?"])

df.shape

vs = df.reindex(np.random.permutation(df.index)) vs.head(10)

vs_names = vs.columns.values vs_names

vs_target = vs.pep

df.dtypes

vs = pd.get_dummies(df) #generating dummy data vs.head(10)

from sklearn.model_selection import train_test_split vs_train, vs_test, vs_target_train, vs_target_test = train_test_split(vs, vs_target, test_size=0.2, random_state=33) print (vs_test.shape) vs_test[0:5]

print (vs_train.shape) vs_train[0:5]

np.set_printoptions(suppress=True, linewidth=120) vs_train.head(10)

vs_test

from sklearn import preprocessing min_max_scaler = preprocessing.MinMaxScaler().fit(vs_train) vs_train_norm = min_max_scaler.transform(vs_train) vs_test_norm = min_max_scaler.transform(vs_test) #preprocessing

np.set_printoptions(precision=2, linewidth=80, suppress=True) #setting precision to 2 vs_train_norm[0:5]

vs_test_norm[0:5]

from sklearn import neighbors, tree, naive_bayes

n_neighbors = 5 #checking for k 5 intially knnclf = neighbors.KNeighborsClassifier(n_neighbors, weights='distance') knnclf.fit(vs_train_norm, vs_target_train)

knnpreds_test = knnclf.predict(vs_test_norm)

print (knnpreds_test)

from sklearn.metrics import classification_report print(classification_report(vs_target_test, knnpreds_test))

from sklearn.metrics import confusion_matrix

knncm = confusion_matrix(vs_target_test, knnpreds_test)

print ("KNN confusion Matrix: ") print (knncm)

print ("KNN Clasification Score test norm: "+str(knnclf.score(vs_test_norm, vs_target_test)))

print ("KNN Clasification Score train norm: "+str(knnclf.score(vs_train_norm, vs_target_train)))

treeclf = tree.DecisionTreeClassifier(criterion='entropy', min_samples_split=3)

treeclf = treeclf.fit(vs_train, vs_target_train)

treepreds_test = treeclf.predict(vs_test) print (treepreds_test)

print (treeclf.score(vs_test, vs_target_test))

print (treeclf.score(vs_train, vs_target_train))

print(classification_report(vs_target_test, treepreds_test))

treecm = confusion_matrix(vs_target_test, treepreds_test, labels=['YES','NO']) print (treecm)

import pylab as plt %matplotlib inline plt.matshow(treecm) plt.title('Confusion matrix') plt.colorbar() plt.ylabel('Actual') plt.xlabel('Predicted') plt.show()

n_neighbors = 11 knnclf = neighbors.KNeighborsClassifier(n_neighbors, weights='distance') knnclf.fit(vs_train_norm, vs_target_train) knnpreds_test = knnclf.predict(vs_test_norm) print(classification_report(vs_target_test, knnpreds_test)) knncm = confusion_matrix(vs_target_test, knnpreds_test) print("Confusion Martix:") print(knncm) print("Score Test") print (knnclf.score(vs_test_norm, vs_target_test)) print("Score Train") print (knnclf.score(vs_train_norm, vs_target_train))

n_neighbors = 11 knnclf = neighbors.KNeighborsClassifier(n_neighbors, weights='uniform') knnclf.fit(vs_train_norm, vs_target_train) knnpreds_test = knnclf.predict(vs_test_norm) print(classification_report(vs_target_test, knnpreds_test)) knncm = confusion_matrix(vs_target_test, knnpreds_test) print("Confusion Martix:") print(knncm) print("Score Test") print (knnclf.score(vs_test_norm, vs_target_test)) print("Score Train") print (knnclf.score(vs_train_norm, vs_target_train))

n_neighbors = 8 knnclf = neighbors.KNeighborsClassifier(n_neighbors, weights='distance') knnclf.fit(vs_train_norm, vs_target_train) knnpreds_test = knnclf.predict(vs_test_norm) print(classification_report(vs_target_test, knnpreds_test)) knncm = confusion_matrix(vs_target_test, knnpreds_test) print("Confusion Martix:") print(knncm) print("Score Test") print (knnclf.score(vs_test_norm, vs_target_test)) print("Score Train") print (knnclf.score(vs_train_norm, vs_target_train))

n_neighbors = 8 knnclf = neighbors.KNeighborsClassifier(n_neighbors, weights='uniform') knnclf.fit(vs_train_norm, vs_target_train) knnpreds_test = knnclf.predict(vs_test_norm) print(classification_report(vs_target_test, knnpreds_test)) knncm = confusion_matrix(vs_target_test, knnpreds_test) print("Confusion Martix:") print(knncm) print("Score Test") print (knnclf.score(vs_test_norm, vs_target_test)) print("Score Train") print (knnclf.score(vs_train_norm, vs_target_train))

n_neighbors =20 knnclf = neighbors.KNeighborsClassifier(n_neighbors, weights='distance') knnclf.fit(vs_train_norm, vs_target_train) knnpreds_test = knnclf.predict(vs_test_norm) print(classification_report(vs_target_test, knnpreds_test)) knncm = confusion_matrix(vs_target_test, knnpreds_test) print("Confusion Martix:") print(knncm) print("Score Test:") print (knnclf.score(vs_test_norm, vs_target_test)) print("Score Train:") print (knnclf.score(vs_train_norm, vs_target_train))

n_neighbors =20 knnclf = neighbors.KNeighborsClassifier(n_neighbors, weights='uniform') knnclf.fit(vs_train_norm, vs_target_train) knnpreds_test = knnclf.predict(vs_test_norm) print(classification_report(vs_target_test, knnpreds_test)) knncm = confusion_matrix(vs_target_test, knnpreds_test) print("Confusion Martix:") print(knncm) print("Score Test:") print (knnclf.score(vs_test_norm, vs_target_test)) print("Score Train:") print (knnclf.score(vs_train_norm, vs_target_train))

treeclf = tree.DecisionTreeClassifier(criterion='entropy', min_samples_split=8) treeclf = treeclf.fit(vs_train, vs_target_train) treepreds_test = treeclf.predict(vs_test) print ("Score Test: "+str(treeclf.score(vs_test, vs_target_test))) print ("Score Train: "+str(treeclf.score(vs_train, vs_target_train))) print(classification_report(vs_target_test, treepreds_test)) treecm = confusion_matrix(vs_target_test, treepreds_test, labels=['YES','NO']) %matplotlib inline plt.matshow(treecm) plt.title('Confusion matrix') plt.colorbar() plt.ylabel('Actual') plt.xlabel('Predicted') plt.show()

treeclf = tree.DecisionTreeClassifier(criterion='entropy', min_samples_split=12) treeclf = treeclf.fit(vs_train, vs_target_train) treepreds_test = treeclf.predict(vs_test) print ("Score Test: "+str(treeclf.score(vs_test, vs_target_test))) print ("Score Train: "+str(treeclf.score(vs_train, vs_target_train))) print(classification_report(vs_target_test, treepreds_test)) treecm = confusion_matrix(vs_target_test, treepreds_test, labels=['YES','NO']) %matplotlib inline plt.matshow(treecm) plt.title('Confusion matrix') plt.colorbar() plt.ylabel('Actual') plt.xlabel('Predicted') plt.show()

treeclf = tree.DecisionTreeClassifier(criterion='entropy', min_samples_split=18) treeclf = treeclf.fit(vs_train, vs_target_train) treepreds_test = treeclf.predict(vs_test) print ("Score Test: "+str(treeclf.score(vs_test, vs_target_test))) print ("Score Train: "+str(treeclf.score(vs_train, vs_target_train))) print(classification_report(vs_target_test, treepreds_test)) treecm = confusion_matrix(vs_target_test, treepreds_test, labels=['YES','NO']) %matplotlib inline plt.matshow(treecm) plt.title('Confusion matrix') plt.colorbar() plt.ylabel('Actual') plt.xlabel('Predicted') plt.show()