Course Coordinator

Name Contact Info
Tailin Wu
wutailin@westlake.edu.cn


Course Interactive Notebooks

https://github.com/AI4Science-WestlakeU/frontiers_in_AI_course


Course Description

This course focuses on the current cutting-edge technologies in the field of computer science and technology, such as Artificial Intelligence, Deep Learning, AI for Science, etc. The lectures are divided into twelve topics, including Frontiers of Deep Learning, Generative Models, Large Models, Reinforcement Learning, Computer Vision and Autonomous Driving, AI + Life Sciences, AI + Scientific Computing, AI + Materials, etc. The content of each topic includes: the development history of theories/technologies, core concepts, underlying ideas and principle mechanisms, the latest research work and technology applications, technology development trends and/or future outlooks, and so on.


Learning Objectives

  1. Have a basic understanding of computer science and technology fields such as artificial intelligence, deep learning, and AI for Science, and understand the research frontiers in the field;
  2. Understand the intersection and important applications of the above tools in various fields;
  3. Conduct in-depth research and study on a field that interests you;
  4. Exercise the ability to read academic papers, write papers, and propose ideas.


Learning Resources

Classic literature in the field of computing, latest high-level academic papers, and influential research by faculty members. The latest papers in prestigious academic journals and international conferences related to the field of computing.


Assessments and Grading

Assessment Criteria Percentage
Attendance
5%
Project proposal and discussion
30%
Project conclusion presentation
30%
Project conclusion report
35%


Course Schedule

Week Theme/Topic Instructor(s) Date Courseware
1
Course Introduction
A general overview of the entire course outline. Give a unified introduction to the fields, importance and frontiers involved in this course from a higher perspective. Introduction to the course paper requirements, grade evaluation, etc. for this class.
Tailin Wu
2024/2/26
01_Course Introduction.pdf
2
Frontiers in Deep Learning
This topic provides an introduction to the theoretical evolution of deep learning and the applications of cutting-edge developments. It first introduces the basic paradigm of deep learning. Next, it introduces the current research frontiers in deep learning, including those in theory, algorithms, and applications.
Tailin Wu
Tao Lin
2024/3/4
02_deep learning fundamentals.pdf
02_optimization.pdf
3
Frontiers in Generative Modeling
This topic introduces the research on generative models centered on diffusion models and generative adversarial networks (GAN). Firstly, it introduces the basic concepts of diffusion models and GANs and the evolution of research, and analyzes the advantages, limitations and development trends of each type of model. We show their wide applications in the fields of image generation (including text-to-image), decision making, and science. Next, the research frontiers in this field are introduced.
Tailin Wu
2024/3/11
03_Frontiers in generative modeling.pdf
4
Foundation models
This topic first introduces the fundamentals of foundation models: the Transformer architecture. Next, it introduces the development history and major branches of big models, including GPT, BERT, etc.. Next, it introduces the development frontiers of foundation models, including multi-modal foundation models, their current limitations, and future directions, etc.
Zhenzhong Lan
2024/3/18
04_Frontiers in foundation model.pdf
5
Introduction to Reinforcement Learning
This topic introduces the fundamentals of reinforcement learning. It first introduces the basic concepts of reinforcement learning and deep reinforcement learning. Next, it introduces the two major branches of deep reinforcement learning: the value function-based approach, and the policy gradient-based approach. Finally, it introduces the research frontiers of reinforcement learning, including offline reinforcement learning, multi-agent reinforcement learning.
Tailin Wu
2024/3/25
05_Introduction to Reinforcement Learning.pdf
6
From AlphaGo to Robotics: Reinforcement Learning Applications
This topic introduces research on cutting-edge applications of deep reinforcement learning in game intelligence and robotics. First, it introduces the applications of reinforcement learning in game intelligence, including DQN for Atari and AlphaGo for Go. Next, it introduces the applications and research frontiers of deep reinforcement learning in robotics.
Donglin Wang
Tailin Wu
2024/4/1
06_Deep Reinforcement Learning and Application to Robotics_Donglin Wang.pdf
06_Reinforcement learning for science_Tailin Wu.pdf
7
Computer Vision and Autonomous Driving
This topic introduces the foundations of computer vision, its main tasks, and the different technologies. Next, the research frontiers of computer vision in autonomous driving are presented in the context of autonomous driving.
Kaicheng Yu
2024/4/8
07_Autonomous Driving_Kaicheng Yu.pdf
8
Course Project Design (1)
Each group project introduces the research problem, importance, difficulties, and major innovations proposed for their project. After the presentation of each project, for the Q&A session with the teacher's comments.
Tailin Wu
2024/4/15
9
Course Project Design (2)
Each group project introduces the research problem, importance, difficulties, and major innovations proposed for their project. After the presentation of each project, for the Q&A session with the teacher's comments.
Tailin Wu
2024/4/22
10
Frontiers in Graph Neural Networks
This topic begins with a brief introduction to the basic concepts of graph neural networks. Next, three important classes of graph neural networks, including GCN, EGNN, and GNS, are introduced, along with their important applications in graph classification, small molecules and proteins, and physics simulation. Finally, the theoretical and applied research frontiers of graph neural networks are presented.
Tailin Wu
2024/4/29
10_Frontiers in Graph neural networks.pdf
11
AI + Life Sciences
This topic will introduce the disciplinary frontiers at the intersection of AI and life sciences. This includes AI for structure prediction, function prediction, and molecular design of small molecules and proteins, and AI for prediction and mechanism discovery in genomes and cells. For the above areas, the main issues in their fields, the main AI approaches, their limitations, and the research frontiers will be presented.
Ziqing Li
2024/5/6
11_AI for Life Science - Selected Topics.pdf
12
Evolutionary Machine Learning and Multi-objective Optimization
This topic will first introduce the basic concepts, main methods, and research frontiers of evolutionary machine learning and multi-objective optimization. Next, it introduces its important applications in the optimization of complex engineering systems, including jet engines, Airbus airframe design, and multi-robot self-organization.
Yaochu Jin
2024/5/13
12_Multi-objective machine learning_Yaochu Jin.pdf
13
AI + Scientific Computing
This topic will introduce the frontiers of the discipline at the intersection of AI and scientific computing. First, the foundations of partial differential equations (PDEs) will be presented, as well as a brief introduction to traditional numerical methods. Next, the basic paradigm of deep learning for modeling partial differential equations will be presented, as well as the main techniques: neural operators, graph-based simulation, ViT, etc. Finally, research frontiers in the field are presented.
Tailin Wu
2024/5/20
13_AI + Scientific Computing.pdf
14
AI + Materials Science
This topic will introduce the disciplinary frontiers at the intersection of AI and materials science. The following aspects of AI for materials will be highlighted: AI-accelerated density functional theory, AI for materials characterization, and materials generation. In each aspect, the concepts, main approaches, and research frontiers will be presented.
Tailin Wu
2024/5/27
14_AI + Materials Science.pdf
15
Course Project Reporting and Discussion (1)
Each group project does a final presentation of the project. After each project presentation, for a Q&A session with instructor critique.
Tailin Wu
2024/6/3
16
Course Project Presentation and Discussion (2)
Each group project does a final presentation of the project. After each project presentation, for a Q&A session with instructor critique.
Tailin Wu
2024/6/11