CS-E4680 - Quantum Machine Learning D, Lecture, 7.9.2023-19.10.2023

Topic outline

• Select topic General

  General
  Welcome to our Seminar on Quantum Machine Learning! In this iteration, we will focus on quantum reinforcement learning (QRL). QRL is a recent and exciting subfield of quantum machine learning that seeks to utilize quantum computing to solve reinforcement learning tasks.

  For this iteration, we decided to have the course presentation-based. That is, every session consists of two presentations each of which is held by two participants. For this, we ask you to team up in pairs using the team selection tool below until this Sunday, 10.9., evening. Some recommended papers are listed below, but feel free to propose your own suggestions.

  Additionally, we want to give you a chance to reflect on every session in a learning diary. We ask you to upload after each session your learnings and reflections on the topics of the week. The deadline for the learning diary is the Sunday evening after the respective session.

  Useful introductory resources

  • To quantum mechanics / quantum computing
    • Online resource: Quantum Country
    • Classic book: Nielsen, M. A., & Chuang, I. L. (2010). Quantum computation and quantum information. Cambridge university press.
  • To reinforcement learning
    • Online resource: A (Long) Peek into Reinforcement Learning
    • Classic book: Sutton, R. S., & Barto, A. G. (2018). Reinforcement learning: An introduction. MIT press.

  Schedule

  Beware that the location changes almost every week! Here is a list with the booked rooms for each date:

  #	Date	Location	Presented papers
  1	Thu, 07.09.23	U414c (Undergraduate Center, Otakaari 1)
  2	Thu, 14.09.23	AS6 (TUAS-building, Maarintie 8)
  3	Thu, 21.09.23	T4 (Computer Science building, Konemiehentie 2)	[4] & [10]
  4	Thu, 28.09.23	M205 (Undergraduate Center, Otakaari 1)	[1] & [3]
  5	Thu, 05.10.23	HILTI (Undergraduate Center, Otakaari 1)	[8] & [2]
  6	Thu, 12.10.23	T4 (Computer Science building, Konemiehentie 2)	[6] & [7]
  7	Thu, 19.10.23	T4 (Computer Science building, Konemiehentie 2)	[5]

  Grading

  The grading will happen based on the presentation. Additionally, attendance and submitting the learning diary are both mandatory. We allow at most one absence (you don't have to submit the learning diary on that day). However, neither attendance nor the learning diary is taken into account for evaluation otherwise.

  Papers

  Here is a list of recommended papers:

  1. [1] Quantum speedup for active learning agents
  2. [2] Quantum Reinforcement Learning
  3. [3] Exponential improvements for quantum-accessible reinforcement learning
  4. [4] Variational quantum circuits for deep reinforcement learning
  5. [5] Quantum enhancements for deep reinforcement learning in large spaces
  6. [6] Quantum reinforcement learning in continuous action space
  7. [7] Variational Quantum Soft Actor-Critic
  8. [8] Parametrized Quantum Policies for Reinforcement Learning
     
  9. [9] Quantum Policy Gradient Algorithm with Optimized Action Decoding
  10. [10] Quantum Reinforcement Learning via Policy Iteration

  • Select activity Topic selection

    

    Topic selection Scheduler
  • Select activity Announcements

    

    Announcements Forum