| | --- |
| | annotations_creators: |
| | - machine-generated |
| | language_creators: |
| | - expert-generated |
| | language: [] |
| | license: |
| | - mit |
| | multilinguality: [] |
| | size_categories: |
| | - 100B<n<1T |
| | source_datasets: |
| | - original |
| | task_categories: |
| | - other |
| | task_ids: [] |
| | pretty_name: How Resilient are Imitation Learning Methods to Sub-Optimal Experts? |
| | tags: |
| | - Imitation Learning |
| | - Expert Trajectories |
| | - Classic Control |
| | --- |
| | |
| | # How Resilient are Imitation Learning Methods to Sub-Optimal Experts? |
| |
|
| | ## Related Work |
| | Trajectories used in [How Resilient are Imitation Learning Methods to Sub-Optimal Experts?]() |
| | The code that uses this data is on GitHub: https://github.com/NathanGavenski/How-resilient-IL-methods-are |
| |
|
| | # Structure |
| | These trajectories are formed by using [Stable Baselines](https://stable-baselines.readthedocs.io/en/master/). |
| | Each file is a dictionary of a set of trajectories with the following keys: |
| |
|
| | * actions: the action in the given timestamp `t` |
| | * obs: current state in the given timestamp `t` |
| | * rewards: reward retrieved after the action in the given timestamp `t` |
| | * episode_returns: The aggregated reward of each episode (each file consists of 5000 runs) |
| | * episode_Starts: Whether that `obs` is the first state of an episode (boolean list) |
| |
|
| | ## Citation Information |
| | ``` |
| | @inproceedings{gavenski2022how, |
| | title={How Resilient are Imitation Learning Methods to Sub-Optimal Experts?}, |
| | author={Nathan Gavenski and Juarez Monteiro and Adilson Medronha and Rodrigo Barros}, |
| | booktitle={2022 Brazilian Conference on Intelligent Systems (BRACIS)}, |
| | year={2022}, |
| | organization={IEEE} |
| | } |
| | ``` |
| |
|
| | ## Contact: |
| | - [Nathan Schneider Gavenski](nathan.gavenski@edu.pucrs.br) |
| | - [Juarez Monteiro](juarez.santos@edu.pucrs.br) |
| | - [Adilson Medronha](adilson.medronha@edu.pucrs.br) |
| | - [Rodrigo C. Barros](rodrigo.barros@pucrs.br) |
| |
|
| |
|
