AR.

Description

Objective: This project focuses on developing an effective model using supervised machine learning algorithms for detecting fraudulent credit card transactions, aiming to minimize false positives and enhance detection accuracy. Dataset: Real-world credit card fraud transactions dataset used for training and evaluation. Methodology: • Data Preprocessing: Utilized pandas for data manipulation, handled missing values, and performed feature scaling with StandardScaler. Data visualization was conducted using matplotlib and seaborn to understand data distributions and correlations. • Machine Learning Algorithms: Employed several supervised learning algorithms: • Logistic Regression • Support Vector Machine (SVM) • Decision Trees • Random Forest • Ensemble Learning Methods (Bagging and Boosting) Model Training and Evaluation: Split data into training and testing sets using train_test_split. Employed GridSearchCV for hyperparameter tuning. Evaluated models using metrics such as accuracy, precision, recall, and F1-score to assess performance. • Performance Metrics: • Accuracy: Overall model correctness. • Precision: Proportion of true positive predictions among all positive predictions. • Recall: Proportion of actual positives correctly predicted. • F1-Score: Harmonic mean of precision and recall, balancing false positives and false negatives. Results: The ensemble learning methods, including Random Forest and XGBoost, demonstrated superior performance over individual algorithms, effectively identifying fraudulent transactions while minimizing false positives.