24-788   Intro to Deep Learning

This course provides a comprehensive introduction to both foundational and advanced topics in deep learning, combining theory, implementation, and real-world applications across diverse engineering domains. Students will begin with the fundamentals of deep neural networks, including feedforward architectures, backpropagation, gradient descent, and regularization, in addition to key model families, such as convolutional, graph, and sequence-based architectures (CNNs, GNNs, LSTMs). Building on this foundation, the course advances to generative modeling and representation learning, covering deterministic and variational autoencoders, GANs, diffusion models, attention mechanisms, including transformers and representative architectures such as BERT, ViT, and GPT. Modern self-supervised learning approaches and foundation models are also explored. Throughout, students will gain hands-on experience through Python-based assignments in PyTorch, applying deep learning concepts to solve real-world problems in mechanical, chemical, biological, electrical, and materials engineering.