Deep Learning Course
“Deep Learning” course (ECE-GY 7123) at NYU Tandon by Chinmay Hegde
The “Deep Learning” course (ECE-GY 7123) at NYU Tandon by Chinmay Hegde offers a deeper dive into the topics learned in the Machine Learning course from Fall 2023 semester. It provides a thorough introduction to neural networks and progresses to cutting-edge topics that define the current DL landscape. This course is ideal for those looking to understand and leverage deep learning in various domains.
Course Overview
The “Deep Learning” course provides a comprehensive exploration of neural network basics, including the foundational principles and architecture. It progresses to neural network training, covering backpropagation, automatic differentiation, and scaling strategies for handling large datasets. Students delve into Convolutional Neural Networks (CNNs) for image processing, Recurrent Neural Networks (RNNs) for sequential data, and attention mechanisms, particularly transformers, for focusing on relevant input data. The course examines large language models like GPT and BERT, integrating deep learning with computer vision for tasks such as object detection and image segmentation. It also covers 3D machine learning models, self-supervised learning techniques, and Generative Adversarial Networks (GANs) for realistic data generation. Additionally, diffusion models are explored for complex data distributions, and Reinforcement Learning (RL) principles are taught, including methods for incorporating human feedback. The course concludes with discussions on sustainability, ethics, and alignment in AI, ensuring a holistic understanding of the broader implications of deep learning.
Course Notes
Notes are available here at professor’s website.
Review
Prof. Chinmay has a remarkable ability to explain complex concepts clearly and effectively, making the subject accessible and engaging. Although the class was conducted online, I can only imagine how enriching the experience would be with in-person discussions and interactions. The assignments, projects, and demos provided by Prof. Chinmay were particularly valuable, offering hands-on experience with deep learning that solidified our understanding and practical skills. Overall, this course was an excellent blend of theory and practice, and I highly recommend it to anyone interested in deep learning.