A-Deep-Dive-into-Imbalanced-Loan-Data-from-lendingclub.com
Dataset Source: https://www.kaggle.com/datasets/itssuru/loan-data The dataset contains 9578 observations and 12 features.
Train test split
The dataset is splited into 80% training set (containing validation set) and 20% test set using stratified sampling.
Exploratory Data Analysis
- Target variable
-
Numerical Feature
- skewness
- different scale
-
Categorical Feature
Feature Engineering
Categorical features:
- One-Hot transformation
Numerical features:
- Feature transformation (log transformation)
- Feature scaling (standardization)
Feature Selection
- Mutual Information
- Recursive Feature Elimination
Dimensionality Reduction
- Principle Component Analysis (PCA)
Sampling
- SMOTE
Models
- Dummy classifier
- Logistic Regression
- Weighted Logistic Regression
- Decision Trees
- Random Forest
- XGBoost
Evaluation Metrics
- Recall
- F1, F2, and F4 scores
Hyperparameter Tuning
- Randomized Search Cross-Validation
- Grid Search Cross-Validation
Final Model Performanceo
============================================================
Model evaluation using K fold stratification
<<< XGBClassifier(base_score=0.5, booster='dart', colsample_bylevel=1,
colsample_bynode=1, colsample_bytree=1, enable_categorical=False,
eta=0.1, gamma=0, gpu_id=-1, importance_type=None,
interaction_constraints='', learning_rate=0.100000001,
max_delta_step=0, max_depth=4, min_child_weight=1, missing=nan,
monotone_constraints='()', n_estimators=200, n_jobs=-1,
num_parallel_tree=1, predictor='auto', random_state=42,
reg_alpha=0, reg_lambda=6, scale_pos_weight=1, subsample=0.8,
tree_method='exact', validate_parameters=1, verbosity=None) >>>
------------------------------------------------------------
mean recall scores, train: 0.8394, test: 0.8041
mean F1 scores, train: 0.8899, test: 0.8507
mean F2 scores, train: 0.8589, test: 0.8221
mean F4 scores, train: 0.8451, test: 0.8093
mean accuracy scores, train: 0.8962, test: 0.8589
============================================================
