💳 Loan Approval Prediction

A Machine Learning Classification Project Predicting loan approval status based on applicant data using machine learning techniques.

**📌 Project Overview ** Loan approval is a critical decision-making process for financial institutions. This project uses machine learning to predict whether a loan will be approved or not based on applicant details such as income, credit history, and loan amount. The objective is to build a robust classification model that helps streamline the loan approval process.

📂 Dataset

Source: Kaggle Description: The dataset includes details such as applicant income, education, marital status, credit history, loan amount, and approval status. Target Variable: Loan_Status (Approved/Not Approved) Link: Kaggle Loan Approval Dataset (https://www.kaggle.com/competitions/playground-series-s4e10)

🛠 Key Features

Exploratory Data Analysis (EDA): Understand data distribution and key patterns. Data Preprocessing: Handle missing values, encode categorical variables, and normalize data. Modeling: Train and test machine learning models for classification. Evaluation: Use performance metrics like accuracy, precision, recall, and F1 score. Insights: Identify factors influencing loan approval decisions.

🔧 Technologies Used

Languages: Python Libraries: Pandas, NumPy, Matplotlib, Seaborn, Scikit-learn Machine Learning Models: Logistic Regression, Decision Trees, Random Forest, XGBoost

📈 Process Workflow

Data Loading: Import the dataset and inspect its structure. EDA: Explore the relationships between features and the target variable. Data Cleaning: Handle missing values and outliers. Feature Engineering: Encode categorical variables and scale numerical features. Model Training: Train classification models such as: Decision Tree Random Forest Lightgbm Model Evaluation: Compare models using metrics like: Accuracy Precision Recall F1 Score Hyperparameter Tuning: Optimize model performance using GridSearchCV. Prediction: Predict loan approval status on test data.

📊 Project Results

Achieved an accuracy of 95.46% on the test dataset. Identified key factors influencing loan approvals, such as credit history and income levels. Improved decision-making with a clear understanding of feature importance.