This repository contains my coursework labs for the Deep Learning course. Each lab focuses on a fundamental concept or technique in deep learning, ranging from basic loss functions to advanced generative models.
-
Lab 1: Mean Square Error (MSE)
Introduction to regression loss functions. Implemented mean square error and explored how it drives optimization in simple models. -
Lab 2: Linear Regression
Building and training a linear regression model with gradient descent. Visualizing the effect of learning rate and convergence. -
Lab 3: Recurrent Neural Networks (RNNs) & Long Short-Term Memory (LSTM)
Sequential data modeling with simple RNNs. Applications discussed include text data and sequence predictions. Implementing LSTMs to address vanishing gradient problems in traditional RNNs. Hands-on experiments with sequence learning tasks. -
Lab 4: Convolutional Neural Networks (CNNs)
Image classification using convolutional architectures. Includes layers such as convolution, pooling, and fully connected layers. -
Lab 5: Transfer Learning
Applying pre-trained models (e.g., ResNet, VGG) on new datasets. Fine-tuning vs feature extraction. -
Lab 6: Machine Translation
Sequence-to-sequence models for translating text between languages using encoder-decoder architectures. -
Lab 7: Generative Adversarial Networks (GANs)
Implementing GANs to generate synthetic data. Training dynamics between generator and discriminator explored.
- Languages: Python
- Frameworks/Libraries: PyTorch / TensorFlow, NumPy, Matplotlib, SciKit-Learn , Keras
- Tools: Jupyter Notebooks