Advanced Data Analysis Project - Lending Club Data
Lending Club:
Prelude) About Lending Club
A) Predictive models (cross-validated efforts)
- EDA / Transformations
- Cross-Validated Model Comparisons
- Model Assumptions / Implications
- Shiny App for users
- Will you get a loan?
- What rate?
B) Who defaults and who doesn't? Inference (classification)
- EDA / Tranformations
- Model Assumptions / Implications
- What variables rise to the top? What does it mean?
- Are they rejecting / accepting the right people?
- Are there any warning signs?
- What would happen if Lending Club changed their practices? (Predictions)
C) How have things changed over time? (Time series)
- Payments / Rates / Defaults / Loan acceptance
- Model fitting / Assumptions
- Have there been any changes in strategy over time?
- What implications will this have in the future if they are maintained?
Others)
- Any exemplary loans that differ from each other?
- Can rates be made differently? Or payment plans?
- Inference: Who gets loans and who doesn't?
Written Report: Due 12/12 @ 5pm: (Double Spaced, Typed, No page limit)
- Introduction / Background
- Objectives (clear story)
- Material & Methods
- Data Source & Descriptions
- Analytical Plans
- Results
- Sensitivity Analysis (to Validate Assumptions)
- Conclusions
- Summary of Findings
- Limitations of Study
- References
- Appendix