Lecture Time:
Tuesday and Thursday, 2:00pm-3:20pm, COA 130
We are taking a hybrid (offline/online) teaching mode, but taking it in-person is more encouraged.
Study Sections:
Office hours instead
TA:
Haiyang Xu (hax038@ucsd.edu)
Zeyuan Chen (zec016@ucsd.edu)
Text Books:
This course is self-contained and we will make the course slides available online, as well as various useful links.
https://d2l.ai/
Office Hours:
Zhuowen Tu, Tuesday (3:30pm-4:30pm, CSB 107, 7:00pm-8:00pm, Zoom), Thursday (3:30pm-4:30pm)
Haiyang Xu, Tuesday (4:30pm-5:30pm, CSB 107), Wednesday (12:30pm-1:30pm, CSB 107)
Zeyuan Chen, Wednesday (4:30pm-5:30pm, Zoom), Thursday (4:30pm-5:30pm, CSB 107)
Deep Learning Resources
Deep Learning (Ian Goodfellow and Yoshua Bengio and Aaron Courville)
Neural Networks for Machine Learning (Geoffrey Hinton)
http://cs231n.stanford.edu/
Course Description:
Recent developments in deep neural networks approaches have led to a significant boost to state-of-the-art learning systems in a wide range of domains including machine learning, robotics, perception, computer vision, artificial intelligence, speech recognition, neural computation, medical imaging, bio-informatics, computational linguistics, and social data analysis. This course will cover the basics about neural networks, as well as recent developments in deep learning including deep belief nets, convolutional neural networks, recurrent neural networks, long-short term memory, and reinforcement learning. We will study details of the deep learning architectures with a focus on learning end-to-end models for these tasks, particularly image classification. Students will learn hands-on skills to do cutting-edge research by implementing, training and debugging neural networks.
Grading policy:
Assignments: 50% (the lowest score out of 5 assignments will be dropped; note that the total points on the HWs will be capped and the bonus credit will only be used to help with the lost points in the assignments)
Midterm: 25%
Final project: 25%
Bonus point: 3% (Piazza, final project)
Late policy: 5% reduction for the first day and 10% reduction afterwards for every extra day past due for the homework assignments and the final project.