A Machine Learning Web App built with Flask
The loan default dataset has 8 variables and 850 records, each record being loan default status for each customer. Each Applicant was rated as “Defaulted” or “Not-Defaulted”. New applicants for loan application can also be evaluated on these 8 predictor variables and classified as a default or non-default based on predictor variables.
bank-loan.csv
There are total 8 attributes which are given below :-
Variable Name - Variable Description - Variable Type
- Age - Age of each customer - Numerical
- Education - Education categories - Categorical
- Employment - Employment status - Numerical
- Address - Geographic area - Numerical
- Income - Gross Income - Numerical
- Debtinc - Individual’s debt - Numerical
- Creddebt - Debt to credit ratio - Numerical
- Othdebt - Any other debts - Numerical
Yes
- Python
- R
- Machine Learning
- Flask
- Spyder
- Jupyter Notebook
- RStudio
- HTML
- CSS
- Logistic Regression
- Decision Tree
- Random Forest
- os
- numpy
- pandas
- pickle
- matplotlib
- seaborn
- sklearn
- Accuracy
- Recall
- Precision
- Specificity
- F1 Score
- AUC - ROC Score
- False Positive Rate
- False Negative Rate
- Bar Graph
- Pie Chart
- Pair Plot
- Box Plot
- Importing libraries
- Loading data set
- Missing value analysis
- Distribution of target variable
- Multicollinearity analysis
- Outlier Analysis
- Missing Values Imputation
- Standardization
- Model training
- Performance metrics
- Model selection
- Freezing best model
