Search Results for author: Edoardo Cetin

Found 9 papers, 2 papers with code

Simple Ingredients for Offline Reinforcement Learning

no code implementations19 Mar 2024 Edoardo Cetin, Andrea Tirinzoni, Matteo Pirotta, Alessandro Lazaric, Yann Ollivier, Ahmed Touati

Offline reinforcement learning algorithms have proven effective on datasets highly connected to the target downstream task.

D4RL reinforcement-learning

A Simple Recipe to Meta-Learn Forward and Backward Transfer

no code implementations ICCV 2023 Edoardo Cetin, Antonio Carta, Oya Celiktutan

Meta-learning holds the potential to provide a general and explicit solution to tackle interference and forgetting in continual learning.

Continual Learning Meta-Learning

Policy Gradient With Serial Markov Chain Reasoning

no code implementations13 Oct 2022 Edoardo Cetin, Oya Celiktutan

We model agent behavior as the steady-state distribution of a parameterized reasoning Markov chain (RMC), optimized with a new tractable estimate of the policy gradient.

Decision Making Reinforcement Learning (RL)

Hyperbolic Deep Reinforcement Learning

no code implementations4 Oct 2022 Edoardo Cetin, Benjamin Chamberlain, Michael Bronstein, Jonathan J Hunt

We propose a new class of deep reinforcement learning (RL) algorithms that model latent representations in hyperbolic space.

Decision Making reinforcement-learning +1

Learning Pessimism for Robust and Efficient Off-Policy Reinforcement Learning

no code implementations7 Oct 2021 Edoardo Cetin, Oya Celiktutan

Off-policy deep reinforcement learning algorithms commonly compensate for overestimation bias during temporal-difference learning by utilizing pessimistic estimates of the expected target returns.

Continuous Control reinforcement-learning +1

Learning Routines for Effective Off-Policy Reinforcement Learning

no code implementations5 Jun 2021 Edoardo Cetin, Oya Celiktutan

Within our framework, agents learn effective behavior over a routine space: a new, higher-level action space, where each routine represents a set of 'equivalent' sequences of granular actions with arbitrary length.

Computational Efficiency reinforcement-learning +1

IB-DRR: Incremental Learning with Information-Back Discrete Representation Replay

no code implementations21 Apr 2021 Jian Jiang, Edoardo Cetin, Oya Celiktutan

However, finding a trade-off between the model performance and the number of samples to save for each class is still an open problem for replay-based incremental learning and is increasingly desirable for real-life applications.

Contrastive Learning Incremental Learning

Domain-Robust Visual Imitation Learning with Mutual Information Constraints

1 code implementation ICLR 2021 Edoardo Cetin, Oya Celiktutan

Human beings are able to understand objectives and learn by simply observing others perform a task.

Imitation Learning

Cannot find the paper you are looking for? You can Submit a new open access paper.