Credit Card Fraud Detection

Table of contents

In this first step you are going to analyze the dataset.

 Is there any Null value in your dataset?
 Display the columns of the dataset.
 Display in percentage the different case(No Frauds, Frauds) present in the dataset.
 Use different colors in other to visualize the distributions of the different case.
 Plot the distribution of of the transaction Amount and transaction time.

Scaling and Distributing

Splitting the Data (Original DataFrame)

Check the Distribution of the labels

Random Under-Sampling: Equally Distributing and Correlating: Correlation Matrices

Anomaly Detection: Dimensionality Reduction and Clustering: implementation and understanding

Classifiers (UnderSampling)

Find the best parameter

Train

Create Dataframe with all score

A Deeper Look into LogisticRegression

SMOTE Technique (Over-Sampling)

Test Data with Logistic Regression

Neural Networks Testing Random UnderSampling Data vs OverSampling (SMOTE)

Keras || OverSampling (SMOTE)