Featured Projects
Cat vs Dog Classification
Binary image classification using both PyTorch and TensorFlow frameworks. Demonstrates transfer learning with VGG16 and MobileNetV2 architectures.
98%
PyTorch Accuracy
98.7%
TensorFlow Accuracy
PyTorch
TensorFlow
VGG16
MobileNetV2
MLflow
Handwritten Digit Recognition
Classic MNIST digit classification using custom CNN architecture with data augmentation techniques for improved generalization.
98%
Test Accuracy
42K
Training Samples
TensorFlow
CNN
Data Augmentation
MNIST
Skin Cancer Classification
Advanced medical image analysis using HAM10000 dataset. Tackles class imbalance with intelligent data augmentation and ResNet-50 transfer learning.
87%
HAM10000 Accuracy
70%
External Validation
PyTorch
ResNet-50
Medical AI
Class Balancing
MLflow