https://github.com/vbhvsingh0/fraudulent_transactions
A few models were developed based on Decision trees and Logistic Regression to categorize fraudulent transactions
https://github.com/vbhvsingh0/fraudulent_transactions
credit-card-fraud data-science logistic-regression machine-learning numpy numpy-python pandas pandas-python python3 random-forest-classifier
Last synced: about 2 months ago
JSON representation
A few models were developed based on Decision trees and Logistic Regression to categorize fraudulent transactions
- Host: GitHub
- URL: https://github.com/vbhvsingh0/fraudulent_transactions
- Owner: vbhvsingh0
- Created: 2024-07-10T02:43:34.000Z (10 months ago)
- Default Branch: main
- Last Pushed: 2024-07-10T04:17:45.000Z (10 months ago)
- Last Synced: 2025-01-14T11:04:31.469Z (4 months ago)
- Topics: credit-card-fraud, data-science, logistic-regression, machine-learning, numpy, numpy-python, pandas, pandas-python, python3, random-forest-classifier
- Language: Python
- Homepage:
- Size: 18.6 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# fraudulent_transactions
A few models were developed based on Decision trees, Logistic Regression, and clustering models to categorize fraudulent transactions
A. Supervised machine learning models
The code inside 'supervised' named folder uses the 'fraud_sampledata.csv' to model the supervised machine learning model to categorize fraud transactions. It has used 3 models:
a. Logistic regression
b. random Forest classifier
c. Ensemble model combining above two.
In terms of performance, the model 'c' was the best having 94 % recall value.
B. Unsupervised machine learning models
The script is present inside the 'unsupervised' named folder. Here, a types of clustering models were used as given below:
a. Kmeans clustering method
b. DBSCAN clustering method
In terms of performance, b worked better suggesting , the shape of the features might be convex shaped.