#importing Libraries

import numpy as np
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import accuracy_score

# loading the dataset

credit_card_data = pd.read_csv('/content/creditcard.csv')

# first 5 rows of the dataset

credit_card_data.head()

Time V1 V2 V3 V4 V5 V6 V7 V8
0 0 -1.359807 -0.072781 2.536347 1.378155 -0.338321 0.462388 0.239599 0.098698

1 0 1.191857 0.266151 0.166480 0.448154 0.060018 -0.082361 -0.078803 0.085102

2 1 -1.358354 -1.340163 1.773209 0.379780 -0.503198 1.800499 0.791461 0.247676

3 1 -0.966272 -0.185226 1.792993 -0.863291 -0.010309 1.247203 0.237609 0.377436

4 2 -1.158233 0.877737 1.548718 0.403034 -0.407193 0.095921 0.592941 -0.270533

5 rows × 31 columns

credit_card_data.tail()

Time V1 V2 V3 V4 V5 V6 V7
15931 27369 -1.160116 -0.244177 0.744250 -0.192350 1.156356 -1.931383 0.409670 -0.3

15932 27369 -3.058318 3.099206 -4.932555 1.924138 -1.576032 -2.135383 -0.830098 2.2

15933 27369 -0.661806 0.315385 2.011194 -0.438757 -0.554990 -0.668072 0.424651 0.0

15934 27370 1.525348 -1.231442 0.420095 -1.551218 -1.376006 0.100758 -1.455755 0.1

15935 27371 1.385680 -0.590076 -0.569197 -0.939441 -0.196015 -0.486685 -0.102496 -0.2

5 rows × 31 columns

# dataset informations
credit_card_data.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 15936 entries, 0 to 15935
Data columns (total 31 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
 0 Time 15936 non-null int64
1 V1 15936 non-null float64
2 V2 15936 non-null float64
3 V3 15936 non-null float64
4 V4 15936 non-null float64
5 V5 15936 non-null float64
6 V6 15936 non-null float64
7 V7 15936 non-null float64
8 V8 15936 non-null float64
9 V9 15936 non-null float64
10 V10 15936 non-null float64
11 V11 15936 non-null float64
12 V12 15936 non-null float64
13 V13 15936 non-null float64
14 V14 15936 non-null float64
15 V15 15936 non-null float64
16 V16 15936 non-null float64
17 V17 15936 non-null float64
18 V18 15936 non-null float64
19 V19 15936 non-null float64
20 V20 15936 non-null float64
21 V21 15936 non-null float64
22 V22 15936 non-null float64
23 V23 15935 non-null float64
24 V24 15935 non-null float64
25 V25 15935 non-null float64
26 V26 15935 non-null float64
27 V27 15935 non-null float64
28 V28 15935 non-null float64
29 Amount 15935 non-null float64
30 Class 15935 non-null float64
dtypes: float64(30), int64(1)
memory usage: 3.8 MB

# checking the missing values in each column

credit_card_data.isnull().sum()

Time 0
V1 0
V2 0
V3 0
V4 0
V5 0
V6 0
V7 0
V8 0
V9 0
V10 0
V11 0
V12 0
V13 0
V14 0
V15 0
V16 0
V17 0
 V18 0
V19 0
V20 0
V21 0
V22 0
V23 1
V24 1
V25 1
V26 1
V27 1
V28 1
Amount 1
Class 1
dtype: int64

# distribution of legit transactions & fraudulent transactions

credit_card_data['Class'].value_counts()

0.0 15862
1.0 73
Name: Class, dtype: int64

# separating the data for analysis

legit = credit_card_data[credit_card_data.Class == 0]
fraud = credit_card_data[credit_card_data.Class == 1]

print(legit.shape)
print(fraud.shape)

(15862, 31)
(73, 31)

# statistical measures of the data

legit.Amount.describe()

count 15862.000000
mean 66.280151
std 188.898885
min 0.000000
25% 5.522500
50% 15.950000
75% 53.890000
max 7712.430000
Name: Amount, dtype: float64

fraud.Amount.describe()

count 73.000000
mean 90.307123
std 271.634360
min 0.000000
25% 1.000000
50% 1.000000
 75% 99.990000
max 1809.680000
Name: Amount, dtype: float64

# comparinge the values for both transactions

credit_card_data.groupby('Class').mean()

Time V1 V2 V3 V4 V5 V6 V
Class
0.0 12104.432165 -0.219072 0.25000 0.862854 0.272641 -0.105868 0.124522 -0.11268

1.0 15559.643836 -7.929807 6.19312 -11.997831 6.555050 -5.474984 -2.480356 -8.35431

2 rows × 30 columns

legit_sample = legit.sample(n=492)

new_dataset = pd.concat([legit_sample, fraud], axis=0)

new_dataset.head()

Time V1 V2 V3 V4 V5 V6 V7
117 76 -1.024576 0.522289 1.787699 0.202672 -1.140803 -0.137831 -0.336555 0.6

8137 11032 1.302043 -0.261503 -0.722756 -0.853728 1.700463 3.426329 -1.013726 0.7

12944 22750 1.244831 -0.323488 0.669388 -0.759828 -0.533733 0.174007 -0.728925 0.0

15403 26790 -0.510225 0.181672 1.449477 -1.939204 0.559337 -0.697591 0.982108 -0.5

1056 803 1.152021 -0.404516 0.864618 -0.026721 -0.880554 -0.112351 -0.545993 0.0

5 rows × 31 columns

new_dataset.tail()
 Time V1 V2 V3 V4 V5 V6 V7
15566 26961 -23.237920 13.487386 -25.188773 6.261733 -17.345188 -4.534989 -17.100492

15736 27163 -23.914101 13.765942 -25.733734 6.290918 -17.784824 -4.572498 -17.390764

15751 27187 -24.590245 14.044567 -26.278701 6.320089 -18.224513 -4.609968 -17.681003

15781 27219 -25.266355 14.323254 -26.823673 6.349248 -18.664251 -4.647403 -17.971212

15810 27252 -25.942434 14.601998 -27.368650 6.378395 -19.104033 -4.684806 -18.261393

5 rows × 31 columns

new_dataset['Class'].value_counts()

0.0 492
1.0 73
Name: Class, dtype: int64

new_dataset.groupby('Class').mean()

Time V1 V2 V3 V4 V5 V6
Class
0.0 11827.691057 -0.227964 0.316819 0.862230 0.250302 -0.053900 0.165045 -0.1098

1.0 15559.643836 -7.929807 6.193120 -11.997831 6.555050 -5.474984 -2.480356 -8.3543

2 rows × 30 columns

X = new_dataset.drop(columns='Class', axis=1)
Y = new_dataset['Class']

print(X)

Time V1 V2 V3 V4 V5 V6 \
117 76 -1.024576 0.522289 1.787699 0.202672 -1.140803 -0.137831
8137 11032 1.302043 -0.261503 -0.722756 -0.853728 1.700463 3.426329
12944 22750 1.244831 -0.323488 0.669388 -0.759828 -0.533733 0.174007
15403 26790 -0.510225 0.181672 1.449477 -1.939204 0.559337 -0.697591
1056 803 1.152021 -0.404516 0.864618 -0.026721 -0.880554 -0.112351
... ... ... ... ... ... ... ...
15566 26961 -23.237920 13.487386 -25.188773 6.261733 -17.345188 -4.534989
15736 27163 -23.914101 13.765942 -25.733734 6.290918 -17.784824 -4.572498
15751 27187 -24.590245 14.044567 -26.278701 6.320089 -18.224513 -4.609968
15781 27219 -25.266355 14.323254 -26.823673 6.349248 -18.664251 -4.647403
15810 27252 -25.942434 14.601998 -27.368650 6.378395 -19.104033 -4.684806

V7 V8 V9 ... V20 V21 V22 \

 117 -0.336555 0.670704 0.071670 ... -0.278966 0.315868 0.847565
8137 -1.013726 0.793970 1.596136 ... 0.046663 -0.119160 -0.231570
12944 -0.728925 0.059987 3.121899 ... -0.136351 -0.180848 0.003462
15403 0.982108 -0.572913 -1.737282 ... 0.146398 0.274250 1.008818
1056 -0.545993 0.046078 0.823252 ... 0.026151 0.074225 0.452292
... ... ... ... ... ... ... ...
15566 -17.100492 15.374630 -3.845567 ... 1.630787 1.769708 -1.691973
15736 -17.390764 15.794136 -3.819832 ... 1.640037 1.773399 -1.748426
15751 -17.681003 16.213627 -3.794093 ... 1.649202 1.777063 -1.804874
15781 -17.971212 16.633103 -3.768351 ... 1.658289 1.780701 -1.861318
15810 -18.261393 17.052566 -3.742605 ... 1.667304 1.784316 -1.917759

V23 V24 V25 V26 V27 V28 Amount

117 0.148877 0.549791 -0.585131 0.325841 -0.068871 0.059713 50.00
8137 -0.051550 1.022013 0.368303 1.046159 -0.086748 -0.000809 15.95
12944 -0.111213 -0.809456 0.520126 -0.644821 0.081947 0.019252 11.85
15403 -0.432498 0.277243 0.255431 -0.306282 -0.381809 -0.381949 15.00
1056 -0.061972 0.226044 0.218102 1.556236 -0.055841 0.006209 34.90
... ... ... ... ... ... ... ...
15566 -1.045673 0.143386 1.611577 -0.221576 1.481233 0.438125 99.99
15736 -1.093104 0.147807 1.663792 -0.221024 1.458076 0.430315 99.99
15751 -1.140605 0.152234 1.715997 -0.220471 1.434951 0.422492 99.99
15781 -1.188167 0.156667 1.768192 -0.219916 1.411855 0.414656 99.99
15810 -1.235787 0.161105 1.820378 -0.219359 1.388786 0.406810 99.99

[565 rows x 30 columns]

print(Y)

117 0.0
8137 0.0
12944 0.0
15403 0.0
1056 0.0
...
15566 1.0
15736 1.0
15751 1.0
15781 1.0
15810 1.0
Name: Class, Length: 565, dtype: float64

#Splitting test and train methods

X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.2, stratify=Y, random_

print(X.shape, X_train.shape, X_test.shape)

(565, 30) (452, 30) (113, 30)

Model Training

Logistic Regression
 model = LogisticRegression()

# training the Logistic Regression Model with Training Data

model.fit(X_train, Y_train)

/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_logistic.py:458: Co
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.

Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regressio
n_iter_i = _check_optimize_result(
▾ LogisticRegression
LogisticRegression()

Model Evaluation

Accuracy Score

# accuracy on training data

X_train_prediction = model.predict(X_train)
training_data_accuracy = accuracy_score(X_train_prediction, Y_train)

print('Accuracy on Training data : ', training_data_accuracy)

Accuracy on Training data : 0.9823008849557522

# accuracy on test data

X_test_prediction = model.predict(X_test)
test_data_accuracy = accuracy_score(X_test_prediction, Y_test)

print('Accuracy score on Test Data : ', test_data_accuracy)

Accuracy score on Test Data : 0.9734513274336283

Could not connect to the reCAPTCHA service. Please check your internet connection and reload to get a reCAPTCHA
challenge.