home_credit

Credit Default Risk — Work Project

This project demonstrates an end-to-end workflow for predicting the probability that an applicant will default on a loan. It includes the problem framing, exploratory analysis, reproducible modeling, and a stakeholder-facing presentation.


Business Problem Statement

Goal. Enable more confident credit decisions by estimating the likelihood of default for new applicants.

Why it matters. Accurate risk estimates help widen access to credit while keeping portfolio risk within target bounds.

Success criteria.

Scope. Supervised classification using application and bureau features; model monitored post-deployment with drift checks.


Exploratory Data Analysis (EDA)

What’s covered.

Key takeaways.


Modeling Workflow (R Markdown)

Approach.

Outputs.

Run this file in RStudio to reproduce the analysis and knit artifacts.


Presentation & Findings

Audience. Product, risk, and operations stakeholders.

Highlights.


How to Navigate

  1. Start with the Business Problem Statement for context and KPIs.
  2. Open the EDA to understand the data and early signal.
  3. Review the Modeling Workflow to see the training/validation approach.
  4. Use the Presentation for an executive summary of results and decisions.