Seminar on Machine Teaching
Saarland University — Winter Semester 2019
This course is offered as a block seminar. The course will provide an overview of machine teaching and cover the following three aspects: (i) how machine teaching formulation differs from machine learning, (ii) the problem space of machine teaching, and (iii) recent work on developing teaching algorithms for human learners.


Course registration at UdS: Timeline

Please check the course registration timeline below depending on your current status (#1, #2):
  1. You received a confirmation from us: If you have already received a confirmation email from us to register, you can go ahead and officially register for the course at UdS. The official registration should start on 16th October, and you will have three weeks of time to officially register.
    Important Note: This is a block seminar, and it won't be possible to withdraw after three weeks of time, i.e., after 5th November. The list of papers and a detailed schedule is provided below. If you have any specific questions regarding the course structure or content, send an email to the instructor. It is important that you make an informed decision to officially register for the course as withdrawal from a seminar course after three weeks is not possible and is marked with a failed grade.
  2. You didn't contact us yet: If you are interested in the course, you can send us an email with the motivation letter as mentioned below in the requirements. We will let you know if there are any slots left in the seminar.


  • Introductory lectures: We will have two introductory lectures on 4th and 5th November at 16:00 — 18:00 in MPI-SWS building E1.5, Room 029. Attendance is mandatory.
  • Research papers: We have assigned a total of 8 research papers to read; list is provided below. You will have to write a two-page report for each paper (one paper per week). Reports for the first 4 papers will be due by 10th December 2019, and reports for the remaining 4 papers will be due by 10th January 2020. A template will be provided in the introductory lecture.
  • Slides preparation: You will have to prepare a presentation of 15 mins for one of the 8 research papers. Each student will be randomly assigned one of the papers to present and this assignment will be done on 10th January 2020. Your final slides will be due by 10th February 2020. You will have a possibility to get feedback on your slides before final submission.
  • Final presentations: At the end of the semester, you will give a final presentation. We will block about 6 hours for the presentations sometime between mid-February to mid-March 2020. The exact dates will be finalized in discussion with enrolled students. Attendance to the final presentations will be mandatory.

List of research papers

  1. On the Complexity of Teaching
    S. Goldman, M. Kearns
    In Proc. of the 4th Conference on Computational Learning Theory (COLT'91), 1991
  2. Deep Knowledge Tracing
    C. Piech, J. Bassen, J. Huang, S. Ganguli, M. Sahami, L. Guibas, J. Sohl-Dickstein
    In Proc. of the 29th Conference on Neural Information Processing Systems (NeurIPS'15), 2015
  3. Learning to Represent Student Knowledge on Programming Exercises Using Deep Learning
    L. Wang, A. Sy, L. Liu, C. Piech
    In Proc. of the 10th International Conference on Educational Data Mining (EDM'17), 2017
  4. Understanding the Role of Adaptivity in Machine Teaching: The Case of Version Space Learners
    Y. Chen, A. Singla, O. Mac Aodha, P. Perona, Y. Yue
    In Proc. of the 32nd Conference on Neural Information Processing Systems (NeurIPS'18), 2018
  5. Learning to Teach in Cooperative Multiagent Reinforcement Learning
    S. Omidshafiei, D. Kim, M. Liu, G. Tesauro, M. Riemer, C. Amato, M. Campbell, J. How
    In Proc. of the 33rd AAAI Conference on Artificial Intelligence (AAAI'19), 2019
  6. Interactive Teaching Algorithms for Inverse Reinforcement Learning
    P. Kamalaruban, R. Devidze, V. Cevher, A. Singla
    In Proc. of the 28th International Joint Conference on Artificial Intelligence (IJCAI'19), 2019
  7. Adversarial attacks on stochastic bandits
    K. Jun, L. Li, Y. Ma, X. Zhu
    In Proc. of the 32nd Conference on Neural Information Processing Systems (NeurIPS'18), 2018
  8. Policy Poisoning in Batch Reinforcement Learning and Control
    Y. Ma, X. Zhang, W. Sun, X. Zhu
    In Proc. of the 33rd Conference on Neural Information Processing Systems (NeurIPS'19), 2019

Course registration at UdS: Requirements

The seminar has a limited number of 12 spots. There are no formal requirements, however, please note the following points:
  • Given limited spots, preference will be given to students who have already taken courses covering topics such as machine learning, data mining, human-computer interaction, statistical learning theory, and optimization.
  • Given limited spots, we are requesting you to provide a short motivation letter (about 5-10 lines) explaining the reasons why you are interested in taking this seminar. Please also mention the relevant courses that you have taken, including your grades. Please send this information via an email to the instructor Dr. Adish Singla (

Imprint / Data Protection