A curated archive of earlier hands-on data-science work across fraud, credit risk, customer analytics, insurance, forecasting, and supervised learning.
Explore · model · validate · interpret
This repository preserves foundational projects from an earlier stage of the portfolio. It shows breadth across classical machine learning, statistical modeling, business analytics, credit/fraud risk, and predictive modeling.
For newer production-style work in security ML, agent security, AI evaluation, detection engineering, graph analytics, and Trust & Safety, see the main GitHub profile.
| Project | Domain | What it demonstrates |
|---|---|---|
| AML — fraud detection using deep learning | Fraud / AML | Predictive modeling for suspicious activity |
| LendingClub | Credit risk | Borrower / default-risk analysis |
| Credit risk default SAS code | Credit risk | SAS-based risk modeling workflow |
| Credit risk model in R | Credit risk | Statistical risk modeling in R |
| Customer lifetime value | Customer analytics | Customer-value estimation and segmentation thinking |
| Insurance claim prediction | Insurance | Claims prediction and supervised ML |
| Loan approval prediction | Lending | Classification for approval decisions |
| Employee attrition | People analytics | Attrition modeling and feature analysis |
| Propensity modeling | Marketing | Customer propensity / response modeling |
| House prices | Regression | Feature engineering and price prediction |
Data preparation
↓
Exploratory analysis
↓
Feature engineering
↓
Classification / regression
↓
Model evaluation
↓
Business interpretation
Techniques: classification · regression · deep learning · risk modeling · customer analytics · feature engineering · EDA
Tools: Python · Jupyter · R · SAS
These notebooks are retained as a foundational archive, so some files may depend on older package versions, local datasets, or notebook-era assumptions. The repository is intentionally presented as historical project work rather than as a modern production package.
For the most current engineering standards—tests, APIs, dashboards, Docker, CI, synthetic evaluation fixtures, model monitoring, and explicit safety boundaries—start with the dedicated flagship repositories on the profile.
Foundations in statistics and ML → production security and AI systems.