Download PDFOpen PDF in browser

Improving Behavioural Cloning with Human-Driven Dynamic Dataset Augmentation

EasyChair Preprint 7363

7 pagesDate: January 21, 2022

Abstract

Behavioural cloning has been extensively used to train agents and is recognized as a fast and solid approach to teach general behaviours based on expert trajectories. Such method follows the supervised learning paradigm and it strongly depends on the distribution of the data. In our paper, we show how combining behavioural cloning with human-in-the-loop training solves some of its flaws and provides an agent task-specific corrections to overcome tricky situations while speeding up the training time and lowering the required resources. To do this, we introduce a novel approach that allows an expert to take control of the agent at any moment during a simulation and provide optimal solutions to its problematic situations. Our experiments show that this approach leads to better policies both in terms of quantitative evaluation and in human-likeliness.

Keyphrases: Behavioural Cloning, Inverse Reinforcement Learning, Reinforcement Learning, hg dagger algorithm, human feedback, human-in-the-loop, machine learning

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
@booklet{EasyChair:7363,
  author    = {Federico Malato and Joona Jehkonen and Ville Hautamaki},
  title     = {Improving Behavioural Cloning with Human-Driven Dynamic Dataset Augmentation},
  howpublished = {EasyChair Preprint 7363},
  year      = {EasyChair, 2022}}
Download PDFOpen PDF in browser