Naive Bayes (BernoulliNB)¶
Credit Card Fraud Detection¶
- Credit card companies need to have the ability to recognize fraudulent credit card transactions so that customers are not charged for items that they did not purchase.
- Datasets contains transactions made by credit cards in September 2013 by european cardholders. This dataset presents transactions that occurred in two days, where we have 492 frauds out of 284,807 transactions. The dataset is highly unbalanced, the positive class (frauds) account for 0.172% of all transactions.
- The data contains only numerical input variables which are the result of a PCA transformation. Unfortunately, due to confidentiality issues, we cannot provide the original features and more background information about the data.
- Input Features: V1, V2, ... V28 are the principal components obtained with PCA, the only features which have not been transformed with PCA are 'Time' and 'Amount'. Feature 'Time' contains the seconds elapsed between each transaction and the first transaction in the dataset. The feature 'Amount' is the transaction Amount, this feature can be used for example-dependant cost-senstive learning.
- Output: 1 in case of fraud and 0 otherwise.
Dr. Ryan @STEMplicity
Importing the Relevant Libraries¶
In [1]:
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
sns.set()
Importing the Dataset¶
In [2]:
df = pd.read_csv('creditcard.csv')
df.head()
Out[2]:
Number & Percentage of Fraud/Not Fraud¶
In [3]:
fraud = df[df['Class'] == 1]
not_fraud = df[df['Class'] == 0]
print("Total =", len(df))
print("\nFraud =", len(fraud))
print("Percentage of Fraud = {:.2f} %".format(1.*len(fraud)/len(df)*100.0))
print("\nNot Fraud =", len(not_fraud))
print("Percentage of Not Fraud = {:.2f} %".format(1.*len(not_fraud)/len(df)*100.0))
Countplot (Fraud/Not Fraud)¶
In [4]:
sns.countplot(df['Class'], palette= 'Set1')
plt.show()
Heatmap (Relationship between Variables)¶
In [5]:
plt.figure(figsize=(40,30))
corr = df.corr()
matrix = np.triu(corr)
sns.heatmap(corr, annot=True, mask=matrix)
plt.show()
Kkdeplot¶
In [6]:
column_headers = df.columns.values
i = 1
fig, ax = plt.subplots(8,4,figsize=(18,30))
for column_header in column_headers:
plt.subplot(8,4,i)
sns.kdeplot(fraud[column_header], bw = 0.4, label = "Fraud", shade=True, color="r", linestyle="--")
sns.kdeplot(not_fraud[column_header], bw = 0.4, label = "Non Fraud", shade=True, color= "y", linestyle=":")
plt.title(column_header, fontsize=12)
i = i + 1
plt.show();
Declaring the Dependent & the Independent Variables¶
In [7]:
X = df.iloc[:, :-1].values
y = df.iloc[:, -1].values
Splitting the Dataset into the Training Set and Test Set¶
In [8]:
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.20, random_state = 7)
Feature Scaling¶
In [9]:
from sklearn.preprocessing import StandardScaler
sc = StandardScaler()
X_train = sc.fit_transform(X_train)
X_test = sc.transform(X_test)
Training the Naive Bayes Model (BernoulliNB)¶
In [10]:
from sklearn.naive_bayes import BernoulliNB
model = BernoulliNB()
model.fit(X_train, y_train)
Out[10]:
Predicting the Test Set Results¶
In [11]:
y_pred = model.predict(X_test)
Confusion Matrix¶
In [12]:
from sklearn.metrics import confusion_matrix, accuracy_score
cm = confusion_matrix(y_test, y_pred)
accuracy = accuracy_score(y_test, y_pred)
print("Accuracy is: {:.2f} %".format(accuracy*100))
sns.heatmap(cm, annot=True, fmt='d')
plt.show()
Classification Report¶
In [13]:
from sklearn.metrics import classification_report
print(classification_report(y_test, y_pred))
K-Fold Cross Validation¶
In [14]:
from sklearn.model_selection import cross_val_score
accuracies = cross_val_score(estimator = model, X = X_train, y = y_train, cv = 10)
print("Accuracy: {:.2f} %".format(accuracies.mean()*100))
print("Standard Deviation: {:.2f} %".format(accuracies.std()*100))
Improving the Model¶
Remove Similar Features¶
In [15]:
df2 = df.drop(['Time','V8','V13','V15','V20','V22','V23','V24','V25','V26','V27','V28'], axis = 1)
df2.head()
Out[15]:
Declaring the Dependent & the Independent Variables¶
In [16]:
X = df2.iloc[:, :-1].values
y = df2.iloc[:, -1].values
Splitting the Dataset into the Training Set and Test Set¶
In [17]:
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.20, random_state =7)
Feature Scaling¶
In [18]:
from sklearn.preprocessing import StandardScaler
sc = StandardScaler()
X_train = sc.fit_transform(X_train)
X_test = sc.transform(X_test)
Training the Naive Bayes Model (BernoulliNB)¶
In [19]:
from sklearn.naive_bayes import BernoulliNB
model = BernoulliNB()
model.fit(X_train, y_train)
Out[19]:
Predicting the Test Set Results¶
In [20]:
y_pred = model.predict(X_test)
Confusion Matrix¶
In [21]:
from sklearn.metrics import confusion_matrix, accuracy_score
cm = confusion_matrix(y_test, y_pred)
accuracy = accuracy_score(y_test, y_pred)
print("Accuracy is: {:.2f} %".format(accuracy*100))
sns.heatmap(cm, annot=True, fmt='d')
plt.show()
Classification Report¶
In [22]:
from sklearn.metrics import classification_report
print(classification_report(y_test, y_pred))
K-Fold Cross Validation¶
In [23]:
from sklearn.model_selection import cross_val_score
accuracies = cross_val_score(estimator = model, X = X_train, y = y_train, cv = 10)
print("Accuracy: {:.2f} %".format(accuracies.mean()*100))
print("Standard Deviation: {:.2f} %".format(accuracies.std()*100))