|Lectures:||Tu, Th 2:00 - 3:15 pm in CAS 216|
|Labs:||Fri 11:15 am-12:05 pm in CGS 521 12:20-1:10 pm in CGS 515|
|Instructor:||Prof. Margrit Betke|
|Teaching Fellow:||Stan (Sha) Lai|
|Class web page:
|Gradescope entry code:
|Staff||Best Way to Contact||Office Hours|
|Margrit Betkeemail@example.com|| Mon 4-6 pm, https://bostonu.zoom.us/my/margritbetke
Tue, Thu 3:15 in CAS 216
Piazza. But note there is a 24-hour timer on your Piazza questions
to encourage other students to answer questions before the TF will able to see
Only send an email if your question is urgent and requires an answer earlier than 24 hours. In that case, send an email to Stan: firstname.lastname@example.org
|Wed 1-3 pm, CDS 824 (or public space nearby)|
Seeing Me on Zoom:
Please feel free to stop by my zoom office hours. I'm happy to talk with you about the course, research in AI, computer vision, machine learning, your plans for the future, or anything else. Check out my personal web page (in need of updating...) to get to know me a little.
Responsibilities of the Teaching Fellows:
Stan is responsible for teaching the two laboratory sections and helping you out during his office hours. He will also help me design the written homework and programming projects, and he will manage the graders. Please contact the TF first if you have questions about your homework grades.
Our goal is to learn about computer systems that exhibit intelligent behavior, in particular, perceptual, reasoning, and robotic systems. Topics include human computer interfaces, computer vision, robotics, machine learning, game playing, knowledge representation, planning, pattern recognition, and natural language processing.
Prerequisites: Calculus, Linear Algebra, Discrete Mathematics, Coding (Python, Java, or C++), Datastructures.
The syllabus below will be updated throughout the semester with lecture notes and reading materials. Links to homework assignments will come alive when the assignment is ready for you to work on. The first assignment will give instructions how you can submit your solution using Gradescope.
Textbook: You are not required to buy a text book. For background reading, you may want to check out the following two books from which I will use materials: Artificial Intelligence by Patrick H. Winston Amazon) and Artificial Intelligence A Modern Approach by Russell and Nordiv, Edition 2 or Edition 3. Both books provide a fundamental background of AI. I will follow, for example, the chapters about Expert Systems and Neural Nets in the Winston book and about Robotics in the Russell and Norvig book, and then augment the materials with newer articles, especially with regards to deep learning.
Class Engagement: Come to class and participate regularly. Reading the assigned texts and listening in lectures and labs will only give you a "passive understanding" of the material. We encourage discussions during lectures and labs to help you acquire an "active understanding" of the material so that you can evaluate existing computer systems critically and learn to develop your own creative solutions. Another way of interacting with the instructors and your classmates is answer students' questions on Piazza.
Homework: Guidelines for submission will be provided with each assignment. We allow teams of two students for the programming assignments. Each student must prepare their own written homework. Late solutions will be levied a late penalty of 20% per day (up to three days). After three days, no credit will be given.
Course Team Project:
The project involves developing a
computer vision system in a team.
https://www.barnesandnoble.com/w/books/1100056731?ean=9780136042594More details will be announced shortly. We will
discuss your project's scope, design, and presentation with you in class, labs,
and office hours. The project has three due dates: (1) your plan, (2) your
initial results, and (3) your final results (see schedule).
Please read the instructions for the course project carefully: Project Instructions (link will become active soon)
Colloquia: You are encouraged to attend the weekly seminar series, Wednesdays, 2-3 pm, of the BU AI Research Initiative. You may also check out AI talks in other departments and at other local universities. A partial list includes: BU College of Engineering Seminar Calendar, CSAIL lab at MIT, Northeastern College of Computer and Information Science, Tufts Department of Computer Science Colloquium.
Talk Reports: You must attend at least three talks on
subjects related to Artificial Intelligence and write a summary on each talk.
Your writeup should give a problem definition, summarize the methods and
results, discuss the work critically, and also briefly explain how the work
relates to material discussed in class. One page is sufficient, do not submit more than two pages.
You may not work in a team but produce your own independent writeup.
Check your text for typographical and grammatical errors, especially if English is not your native language. Use tools to check for error, e.g., grammarly.com, ChatGPT. You must acknowledge use of these tools. You will lose points if you simply copy the speaker's abstract, your review is late, contains typographical and grammatical errors, does not contain a discussion, and does not contain a statement how the talk relates to class material.
Exams: There will be two exams on the material discussed in the class and practiced with homeworks. The exams will be quite easy for students who attend class, participate in our discussions, and keep up with homework assignments and programming projects. The midterm quiz will be on October 5, 2023, final exam will be during the finals period in December and will be determined by the University.
Grading Policy: Your final grade will be determined as follows:
|6 Homework Sets||34%|
|Class Participation/Engagement (Lecture, Lab, Piazza)||10%|
|Team Programming Project and its Presentation||10%|
|3 AI Talk Reports||6%|
You are encouraged to discuss course materials on Piazza and allowed to collaborate with classmates on the solution of the homework. If you do, you must acknowledge your collaborators. Each student must submit his or her own electronic version of the solutions. While the code of collaborating students may be the same, comments explaining the code and answers to exercises should not not be identical but based on the individual student's work. If you use algorithms or code that are not your own original work and that were not provided in class or discussed in the reading materials, or if you use ChatGPT or similar AI tools for help with your solutions, you must give a detailed acknowledgment of your source.
Cheating and plagiarism are not worthy of Boston University students. We expect you to abide by the rule stated above and the standards of academic honesty and computer ethics policy described in http://www.bu.edu/computing/ethics/ and http://www.bu.edu/academics/policies/academic-conduct-code.
Artificial Intelligence is a very exciting topic, and the course should be fun and not too much of a struggle for you. Make sure that you have had the prerequisites. Depending on your level of programming experience and/or mathematics background, the course may be challenging for you. If you do not understand the material, ask for help immediately. Ask questions in class. If one student is confused about something, then maybe others are also confused and grateful that someone asked. Please interact with the TF and me for help. Our task is to help you learn a very interesting topic. Also interact on Piazza with your classmates. This is your chance to build your network of AI colleagues.
|Lecture Dates||Topics and Lecture Materials||Additional Readings||Homework & Talk Reviews|
|2023 Tu 9/5||Introduction: What is AI? Turing Test. Industry Successes||Turing test. Russell: Intro, Intelligent Agents. Self-driving Cars in San Francisco||Talk reviews 1--3 assigned.|
|2023 Th 9/7||
Research in AI at BU.
Measures of Success in AI: ROC analysis, sensitivity & specificity, precision & recall, F1 score, balanced accuracy, confusion matrices.
|Fawcett Wiki on F1 score||
HW1 (ROC, Responsible AI)
|2023 Tu 9/12||
Sources of Harm throughout the Machine Learning Life Cycle.
Racial Bias in Emotion Recognition AI Systems. Carbon Footprint of Large Language Models.
Research in AI at BU.
Rhue: Biased Emotion reading,
Pasquale: ML and facial inference to classify people,
Buolamwini PhD thesis,
Biases through ML Life Cycle,
Exposing Political Orientation from Faces, Stochastic Parrots.
|2023 Th 9/14, Tu 9/19, Th 9/21||
Machine Learning and Neural Nets.
Learning from Examples. Learning by Training Neural Nets -- An Example.
Backpropagation: Partial Derivative Examples, Backprop Derivation , Backprop Derivation Notes
Last day to ADD class: Wednesday, 9/18/2023
Chennuri et al., 2023
Wikipedia 1, 2, 3 Activation functions
H2 (ML, NN)
|2023 Tu 9/26, Th 9/28 (SCC), Tu 10/3||
More about Training A Network
NLP and Transformers
BU's Scientific Computing Cluster (SCC).
|LeCun 1990 paper, Convolutional Nets, Sejnowski and NetTalk.mp3|
|2023 Th 10/5|| Midterm Quiz
|2023 Tu 10/10|| No class. Monday schedule.
Last day to DROP class without a "W": Tuesday, 10/10/2023
|2023 Th 10/12||Computer Vision Introduction. Object Recognizability, Coherence, and Convolution, and Neural Nets: 1st Lecture on Computer Vision.||Russell: Perception.||.|
|2023 Tu 10/17, Thu 10/19||ImageNet. Face Recognition and Other Biometrics||ImageNet website. LFW website. MegaFace. FaceScrub. IARPA Janus Benchmark A, CFP Dataset, AFLW 2000 Dataset. Cao et al. 2018, Zhu et al. 2019 Zhu et al. 2020, Zhu et al. 2022 Best-Rowden, Jain 2017 Banerjee et al., 2013. IAB IIT Jodhpur West Virginia biometrics||
|2023 Tu 10/24, Th 10/26||
Interfaces for people with disabilities:
Multimodal News Analysis.
|.||Computer Vision Project Plan|
|2023 Tu 10/31||Markov Models, Hidden Markov Models with Discrete Observations, Training HMMs||Rabiner (up to p. 275)||Project start|
|2023 Th 11/2, Tu 11/7||
HMMs with Continuous Output
Pathology Image Analysis,
Last day to DROP class with a "W": Monday, 11/13/2023
|2023 Th 11/9, Tu 11/14||Markov Decision Processes and Reinforcement Learning.||Russell 21||.|
|2023 Th 11/16, Tu 11/23||Game Playing: Minimax Alpha-Beta Algorithm for Adversarial Games. Lecture notes on Adversarial Search. Heuristic Pruning of Game Trees (progressive deepening, tapered search, horizon effect, continuation heuristics). Stochastic Games. Partially Observable Games.||
Russell: Adversarial Search (Ch 5),
IBM's Chess Playing Computer versus Kasparov. FIDE, World Chess Federation Network.
Global Games Market 2017, 2018, 2019, 2022. Watson, IBM's Jeopardy! playing computing system,
|2023 Wed 11/22- Sun 11/26||Thanksgiving Recess||.||.|
|2023 Tu 11/28||Search Strategies & Robot Path Planning: Lecture notes.||Lozano-Perez, Robot Motion Planning, Visibility Graph.||.|
|2023 Th 11/30||Rule-based Deduction Systems, Forward and Backward Chaining, Conflict Resolution Strategies.||Russell||H6|
|2023 Tu 12/5||
Logic and Planning,
Resolution Proofs. Situation Variables. Planning
and Acting in the Real World.
||Russell: First-Order Logic, Inference, Classical Planning, Planning and Acting in the Real World (Ch 8, 9, 10.1, 10.4, 11.4). Wikipedia: 1, 2, 3.||H5 due.|
|2023 Th 12/7||Talk presentations by students teams (5 minutes each).||.||640 Final Project Report due|
|2020 Tu 12/12|| Robotics,
Moral Competence of Robots.
Announcing Winner of Game Competition
|Article on Pepper Russell: Robotics (Ch 25), What Makes a Robot Likable? Learning How to Behave: Moral Competence for Social Robots||H6 due. Talk review 1, 2, 3 due|
|2023 December 15-21||Final Exam Day and Location TBA||.||.|
The lecture schedule may change depending on the time spent on each topic and whether alternative subjects are discussed. Suggestions for additional topics are welcome!
Check out http://www.cs.bu.edu/faculty/betke/links.html if you need ideas for your class project, if you are looking for a job, or if you are interested in research related to AI and computer vision. You will find a list of links to conferences, journals, research groups, and companies.