credit-card-fraud-detection

steps

1. performing EDA and writing some observations

2. visualizing data

3. pca,tsne

4. dealing imbalanced data into two type

   1. under sampling
   2. over sampling

1) under sampling

1. removing outliers
2. splitting of data
3. building model 1)logistic regression 2)random forest 3)xg boost 4)decision tress 5)svm 6)grading boost
4. hypertuning of all the model
5. predicting the results

2) over sampling

1. smote over sampling
2. building same models
3. hypertuning
4. predicting the results