• Problem Statement
• Data source
• Steps
• Tech Stack
• Results

This model forecasts how much of a loan an applicant will receive. The model forecasts how much will be permitted based on various variables about the applicant profile. A co-applicant with a higher credit score is usually awarded a larger loan amount. It also depends on how much the applicant has asked for.

• Kaggle loan amount prediction

• Exploratory data analysis
• Bivariate analysis
• Multivariate correlation

• Python (refer to requirement.txt for the packages used in this project)

Top 3 models (with default parameters)

• The final model used is: Random Forest
• Metrics used: RMSE
• Why choose random forest while bagging yields the best results?: Comparing the RMSE while tuning the parameters, random forest produced the lowest RMSE consistently.

• Based on the analysis of this project, we found out that the loan amount that will be granted is determined mainly by the loan amount requested, credit score, and a co-applicant. The least important features are expense types and gender.
• Recommendation would be to focus more on the most predictive feature when looking at the applicant profile and pay less attention to the least predictive features.