celik-muhammed / dl_ann_ml_capstonep-fraud-detection-classification-project Goto Github PK

• The Capstone Project Of Machine Learning And Deep Learning Module
• The datasets contains transactions made by credit cards in September 2013 by european cardholders. This dataset presents transactions that occurred in two days, where it has 492 frauds out of 284,807 transactions. The dataset is highly unbalanced, the positive class (frauds) account for 0.172% of all transactions.
• You will implement Logistic Regression, Random Forest, XGBoost,and Neural Network algorithms and Unbalanced Data Techniques . Also visualize performances of the models using Seaborn, Matplotlib and Yellowbrick in a variety of ways.
• At the end of the project, you will have the opportunity to deploy your model by Streamlit API.

1. DL_CapstoneP Fraud Detection V3 final.ipynb

Visually Compare Models Performance In a Graph

Scores

Random Forest

Feature Importance Computed in 3 Ways with Python

Random Forest

Permutation Based Feature Importance

Importance computed with SHAP values

Data Review

(Visual Method) Create a histogram.

(Visual Method) Create a boxplot.

(Visual Method) Create a Q-Q plot.

Streamlit