back to search

Approximate Dynamic Programming and Reinforcement Learning

EI7649Master's Modules6 ECTSEnglishwinter semesterDepartment Computer Engineering
AI-edited module sheet. Based on the TUMonline module description, edited for readability.Original in TUMonline

What it is about

You will learn methods of Approximate Dynamic Programming (ADP) and Reinforcement Learning (RL) to solve sequential decision problems. In the end you will be able to describe fundamental models and algorithms, follow derivations, and implement simple ADP/RL methods and apply them to simple tasks (e.g., robotic).

What you will be able to do

  • Describe classical scenarios of sequential decision problems
  • Explain basic models of ADP/RL
  • Derive ADP-/RL- algorithms from the course
  • Characterize convergence properties of the algorithms studied
  • Theoretically and practically compare performance differences of the algorithms
  • Select suitable algorithms for concrete applications
  • Construct and implement ADP/RL algorithms

What the module consists of

  • VorlesungDelivery of theoretical foundations and models
  • ÜbungPractice and deepening of concepts; covers sometimes new topics
  • TutoriumDiscussion of exercises and programming tasks; support in their solution

Teaching method

  • Frontalunterricht (Tafel, Beamer)Introduction to theory and models
  • Gruppen- und EinzelbesprechungenDeveloping definitions and concepts using simple examples
  • Übungs- und ProgrammieraufgabenPractice-oriented application of lecture content

Dates

Lecture with exerciseApproximate Dynamic Programming and Reinforcement Learning2 groups to choose from

  • AThu13:15–14:450220, Hörsaal m. Exp.-Bühne (0502.EG.220)
    14× · 15.10.–04.02.
    • 15.10.
    • 22.10.
    • 29.10.
    • 05.11.
    • 12.11.
    • 19.11.
    • 26.11.
    • 10.12.
    • 17.12.
    • 07.01.
    • 14.01.
    • 21.01.
    • 28.01.
    • 04.02.
  • BMon15:00–16:300220, Hörsaal m. Exp.-Bühne (0502.EG.220)
    14× · 19.10.–01.02.
    • 19.10.
    • 26.10.
    • 02.11.
    • 09.11.
    • 16.11.
    • 23.11.
    • 30.11.
    • 07.12.
    • 14.12.
    • 21.12.
    • 11.01.
    • 18.01.
    • 25.01.
    • 01.02.

From the current semester, not binding. You attend one of several groups; the timetable automatically suggests the one with the fewest clashes.

Show TUMonline data
Sprache
Englisch
Turnus
Wintersemester
Modulniveau
Master
Moduldauer
Einsemestrig
Gesamtstunden
180
Präsenzstunden
75
Eigenstudiumstunden
105
Organisationsname
Department Computer Engineering

Courses

  • Approximate Dynamic Programming and Reinforcement Learning
  • Approximate Dynamic Programming and Reinforcement Learning
  • Approximate Dynamic Programming and Reinforcement Learning - Fragestunde

Official page in TUMonline · Details are not binding.