Human-Interactive Robot Learning (HIRL)

Graduate level educational module covering basic principles and techniques of Human-Interactive Robot Learning.

A collaboration between ISIR, Sorbonne University, Vrije Universiteit Amsterdam

What is this module about?

Module context

Machine learning techniques that allow robots to learn from people (mostly based on Reinforcement Learning)
Human-centered (teaching experience, usability, …), thinking about HRI, multi-disciplinary flavor
Combination of lectures and practicals

Goals:

Explain what a Human-Interactive Robot Learning (HIRL) problem is
List a examples of HIRL systems across different domains
Explain the relationship between HIRL, robot learning, interactive machine learning, and classical machine learning
List different types of teaching signals, communication modalities
Connecting HIRL to theories from other fields like social learning theory - scaffolding - curriculum learning

Location of HIRL between Interactive Machine Learning and different learning types

Prerequisites

Experience with Python, Machine Learning and Robotics helps

How is this module designed to be used?

Modular design enables personal choice of how to use the material

Frequently Asked Questions (FAQ):

FAQ

Page updated

Google Sites

Report abuse