Developed a machine learning model to detect online payment fraud. Used Logistic Regression, SVM, and Random Forest, and evaluated performance using various metrics. Visualized data for insights and selected the most effective model.
Data Acquisition and Preprocessing
- Collected a relevant dataset containing historical online payment transactions, including features such as transaction amount, time, location, and customer information.
- Performed data cleaning and preprocessing tasks to handle missing values, outliers, and inconsistencies.
- Feature engineering was conducted to create new or derived features that could improve model performance. Model Selection and Training:
Employed three popular machine learning algorithms
- Logistic Regression, Support Vector Machine (SVM), and Random Forest.
- Each model was trained on the preprocessed dataset, using appropriate hyperparameter tuning techniques to optimize performance.
Model Evaluation
- Evaluated the trained models using relevant performance metrics such as accuracy, precision, recall, F1-score, and confusion matrix.
- Cross-validation techniques were employed to assess model generalization and prevent overfitting.
Visualization and Analysis
- Utilized data visualization techniques to gain insights into the data and model behavior.
- Created visualizations like heatmaps, distribution plots, and count plots to explore relationships between features, class distributions, and model predictions.