Texas A&M UniversityCSCE 625 Artificial Intelligence |
||
Spring 2025 |
||
Description: This graduate-level course will focus on the fundamental concepts and modern methods of artificial intelligence such as search (uninformed, informed, iterative improvement, constraint satisfaction, space/time complexity), game playing (minmax, alpha-beta pruning; knowledge representation and reasoning; propositional logic, first-order logic and automated theorem proving, etc.), planning, uncertainty and probabilistic reasoning, machine learning and deep learning basics. Selected topics include foundation models, generative AI, agentic AI, physical AI, robotics, natural language processing, computer vision, and AI ethics.
** class schedule is subject to change **
AI: Artificial intelligence: a modern approach (3rd edition). Stuart Russel and Peter Norvig
PRML: Pattern recognition and machine learning. Springer, 2006. Christopher M Bishop
Lecture | Date | Topic | Additional Reading | Note | |
Week 1: Overview and AI agent | |||||
1 | Tu 1/14 | Logistics and course overview [slides] | Review: linear algebera, probability, and Python | ||
2 | Th 1/16 | AI agent design [slides] | AI: 1-2 | Drop out due: 1/17 | |
Week 2: Search agent | |||||
Tu 1/23 | |
AI: 3.1-3.5 AI: 5.1-5.3 |
|||
3 & 4 | Th 1/16 | Uninformed & Informed search
[slides]
[slides]
+ detailsBFS, DFC, UCS, Greedy, and A* search |
AI: 3.1-3.5 AI: 5.1-5.3 |
||
Week 3: Search agent | |||||
5 | Tu 1/28 | Graph search [slides] |
AI: 3.1-3.5 AI: 5.1-5.3 |
Assignment 1 out [link] | |
6 | Th 1/30 | Adversarial search [slides] |
AI: 3.1-3.5 AI: 5.1-5.3 |
||
Week 4: Learning agent (concept and data) | |||||
7 | Tu 2/4 | Learning agents and data
[slides]
+ detailsMaching learning overview, application data |
PRML: 1.2.1, 1.2.2, 1.2.4 PRML: 2.5 |
||
8 | Th 2/6 | Data, feature, and representation
[slides]
+ detailsBag of words, histograms |
PRML: 1.2.1, 1.2.2, 1.2.4 PRML: 2.5 |
||
Week 5: Learning agent (feature and representation) | |||||
9 | Tu 2/11 | Correlation and normalization
[slides]
+ detailsNon-parametric representations (Parzen), data correlation, Z-score |
PRML: 1.4 PRML: 12.1, 12.4.3 |
Assignment 1 due Assignment 2 out [link] |
|
10 | Th 2/13 | Dimensionality reduction
[slides]
+ detailsPCA & embedding (T-SNE) |
PRML: 1.4 PRML: 12.1, 12.4.3 |
||
Week 6: Learning agent (optimization basics) | |||||
11 | Tu 2/18 | Linear regression
[slides]
+ detailsGeneral parameter estimation techniques |
PRML: 1.1, 1.2.5, 1.5.5 PRML: 3.1 |
||
12 | Th 2/20 | Course project topic discussion [slides] | Refer to Slack for the recording | ||
Week 7: Learning agent (optimization basics, deep learning) | |||||
13 | Tu 2/25 | Non-linear regression
[slides]
+ detailsGradient descent and Newton's method, optimization basics |
PRML: 1.1, 1.2.5, 1.5.5 PRML: 3.1 |
||
14 | Th 2/27 | Neural nets and deep learning basics
[slides]
+ detailsDeep learning introduction with CNN, RNN, GNN basics, perceptron, multi-layer perceptron, backpropagation, training particulars |
PRML: 4.1.7 PRML: 5.1-5.3 | ||
Week 8: Learning agent (models) | |||||
15 | Tu 3/4 | Self-attention and Transformers [slides] |
Attention Is All You Need RNN and Transformer (Chapter) The Annotated Transformer Neural Machine Translation Vision Transformers |
Assignment 2 due Assignment 3 out [link] | |
Th 3/6 | Project highlight presentation | ||||
Week 9 | |||||
Tu 3/11 | Spring break (no class) | ||||
Th 3/13 | Spring break (no class) | ||||
Week 10: Foundation model basics | |||||
16 | Tu 3/18 | Pre-training [slides]
+ detailsUnsupervised learning: K-means, PCA, Autoencoder. Self-supervised learning: contrastive learning, predictive learning. |
Representation learning (Chapter) Review paper Lil's blog: contrastive learning DINOv2, CLIP, MAE |
||
17 | Th 3/20 | Post-training and adaptation [slides]
+ detailsSupervised fine-tuning (SFT), parameter-efficient transfer learning (PETL), (visual) instruct tuning, re-parameterization, RLHF |
Llama 3 Sebastian's blog Visual prompt tuning Visual prompt Prompt via inpainting Open AI RLHF, o1-preview CoT |
||
Week 11: Generative models | |||||
18 | Tu 3/25 | Deep generative models [slides] |
Generative models (Chapter) Generative modeling meets representation learning (Chapter) Conditional generation (Chapter) |
||
19 | Th 3/27 | Model zoo: diffusion, auto-regressive [slides] |
VAE Lil' blog: diffusion models Yang Song's blog Autoregressive Generation |
||
Week 12: Decision making & agents | |||||
20 | Tu 4/1 | Diffusion models (cont.) [slides] |
VAE Lil' blog: diffusion models Yang Song's blog Autoregressive Generation |
Assignment 3 due | |
21 | Th 4/3 | Reinforcement learning & robot learning [slides] |
Survey FMs for robot learning SayCan RT-1 3D Diffuser Actor |
||
Week 13: Decision making & agents | |||||
22 | Tu 4/8 | World model [slides] | Assignment 4 out [link] | ||
23 | Th 4/10 | 3D Human Foundation Agents [slides] |
Michael Black's talk |
||
Week 14: Decision making & agents | |||||
24 | Tu 4/15 | Language and computer use agents [slides] |
Building effective agents Lil' blog: LLM powered autonomous agents Yu Su's blog: language agents MOOC on LLM agents Computer-using agent |
||
Th 4/17 | Bridging Vision and Sound: Audio-Visual Scene Perception and Generation (Prof. Yapeng Tian) | Guest lecture | |||
Week 15 | |||||
Tu 4/22 | Final presentation | ||||
Th 4/24 | Final presentation | ||||
Week 16 | |||||
Tu 4/29 | Final presentation (if needed) | Assignment 4 due | |||
Th 5/1 | Reading day (no class) |