Performance Metrics for State-Based Imitation Learning

Autor/innen

  • Mohamed Zalat Carleton University
  • Babak Esfandiari

DOI:

https://doi.org/10.32473/flairs.v34i1.128479

Schlagworte:

Imitation Learning, Metrics, Machine Learning, Agent-Based Systems

Abstract

We propose five new domain-independent metrics for evaluating and comparing performance at imitating a state-based expert. We use two agents in the RoboCup environment to compare the performance metrics: an agent based on a Multi-Layer Perceptron (MLP) and an agent based on a Long Short-Term Memory (LSTM) neural network.

Downloads

Veröffentlicht

2021-04-18

Zitationsvorschlag

Zalat, M., & Esfandiari, B. (2021). Performance Metrics for State-Based Imitation Learning. The International FLAIRS Conference Proceedings, 34. https://doi.org/10.32473/flairs.v34i1.128479

Ausgabe

Bd. 34 (2021): Proceedings of FLAIRS-34

Rubrik

Main Track Proceedings