Research-oriented project investigating the Double Descent phenomenon in overparameterized deep neural networks. Conducted experiments on CIFAR-10 using ResNet and Vision Transformer architectures, analyzing interpolation thresholds, generalization behavior, and the effects of implicit regularization under varying model capacities.
🔗 View ProjectDeveloped a machine learning pipeline to predict employee attrition using the IBM HR Analytics dataset. Performed data preprocessing, feature engineering, class imbalance handling with SMOTE, and hyperparameter tuning. Evaluated multiple models using ROC-AUC, precision, recall, and F1-score to identify employees at risk of leaving.
🔗 View ProjectImplemented and analyzed iterative methods for solving linear systems, including Jacobi and Gauss-Seidel algorithms. Compared convergence rates, computational complexity, and parallel implementations to study their performance and scalability on modern hardware architectures.
🔗 View Project