Problem Statement : the number of customers defaults reached 50,968 people and the company suffered a loss of $743,972,450 in 2007 to 2014
Goals : reduce company losses by up to 30 percent of bad loan
Objective : Create a ML system to help loan assessments automatically
Business Metrics : Cost
Result :
Tools: Python, JupyterLab, Git
Libraries: Pandas, Numpy, Feature-engine, Scikit-learn, Imbalanced-learn, statistic-learn, imputer-learn, WoE Binning
Dataset: Loan Data 2007-2014 [source]
Summary of the analysis
- This dataset have 466,285 observations and 74 variables with 52 numerical variables, and 22 categorical variables.
What I have learned
- Framing the business problem.
- Create a machine learning model with optimal of number of approved and number of rejected
- Create a scorecard that can generate credit score who rejected
- Make a business simulation from machine learning model.
File Dictionaries
- Credit_Risk_Model_VIX_IDX_Partners_Archie_Citra_Muhammad.ipynb: this notebook contains all of project details, such as Problem Research, exploratory data analysis & insights from dataset, data preprocessing, modeling, scorecard, business recommendation
- Credit_Risk_Model_VIX_IDX_Partners_Archie_Citra_Muhammad.pdf: summary of the project.