no code implementations • 4 Mar 2024 • Théo Vincent, Daniel Palenicek, Boris Belousov, Jan Peters, Carlo D'Eramo
Value-based Reinforcement Learning (RL) methods rely on the application of the Bellman operator, which needs to be approximated from samples.
1 code implementation • ICLR 2020 • Carlo D'Eramo, Davide Tateo, Andrea Bonarini, Marcello Restelli, Jan Peters
We study the benefit of sharing representations among tasks to enable the effective use of deep neural networks in Multi-Task Reinforcement Learning.
1 code implementation • 20 Dec 2023 • Théo Vincent, Alberto Maria Metelli, Boris Belousov, Jan Peters, Marcello Restelli, Carlo D'Eramo
We formulate an optimization problem to learn PBO for generic sequential decision-making problems, and we theoretically analyze its properties in two representative classes of RL problems.
no code implementations • 5 Dec 2023 • Erdi Sayar, Zhenshan Bing, Carlo D'Eramo, Ozgur S. Oguz, Alois Knoll
Multi-goal robot manipulation tasks with sparse rewards are difficult for reinforcement learning (RL) algorithms due to the inefficiency in collecting successful experiences.
1 code implementation • 19 Nov 2023 • Ahmed Hendawy, Jan Peters, Carlo D'Eramo
Multi-Task Reinforcement Learning (MTRL) tackles the long-standing problem of endowing agents with skills that generalize across a variety of problems.
no code implementations • 3 Nov 2023 • Gabriele Tiboni, Pascal Klink, Jan Peters, Tatiana Tommasi, Carlo D'Eramo, Georgia Chalvatzaki
Varying dynamics parameters in simulation is a popular Domain Randomization (DR) approach for overcoming the reality gap in Reinforcement Learning (RL).
no code implementations • 3 Nov 2023 • Aryaman Reddi, Maximilian Tölle, Jan Peters, Georgia Chalvatzaki, Carlo D'Eramo
To this end, Robust Adversarial Reinforcement Learning (RARL) trains a protagonist against destabilizing forces exercised by an adversary in a competitive zero-sum Markov game, whose optimal solution, i. e., rational strategy, corresponds to a Nash equilibrium.
no code implementations • 25 Sep 2023 • Pascal Klink, Carlo D'Eramo, Jan Peters, Joni Pajarinen
In this work, we focus on framing curricula as interpolations between task distributions, which has previously been shown to be a viable approach to CRL.
no code implementations • 19 Sep 2023 • Tuan Dam, Pascal Stenger, Lukas Schneider, Joni Pajarinen, Carlo D'Eramo, Odalric-Ambrym Maillard
We introduce a novel backup operator that computes value nodes as the Wasserstein barycenter of their action-value children nodes; thus, propagating the uncertainty of the estimate across the tree to the root node.
no code implementations • 11 Feb 2022 • Tuan Dam, Carlo D'Eramo, Jan Peters, Joni Pajarinen
In this work, we propose two methods for improving the convergence rate and exploration based on a newly introduced backup operator and entropy regularization.
no code implementations • ICLR 2022 • Pascal Klink, Carlo D'Eramo, Jan Peters, Joni Pajarinen
This approach, which we refer to as boosted curriculum reinforcement learning (BCRL), has the benefit of naturally increasing the representativeness of the functional space by adding a new residual each time a new task is presented.
no code implementations • 11 May 2021 • Julen Urain, Anqi Li, Puze Liu, Carlo D'Eramo, Jan Peters
Reactive motion generation problems are usually solved by computing actions as a sum of policies.
1 code implementation • 25 Mar 2021 • Andrew S. Morgan, Daljeet Nandha, Georgia Chalvatzaki, Carlo D'Eramo, Aaron M. Dollar, Jan Peters
Substantial advancements to model-based reinforcement learning algorithms have been impeded by the model-bias induced by the collected data, which generally hurts performance.
Model-based Reinforcement Learning Model Predictive Control +2
1 code implementation • 25 Feb 2021 • Pascal Klink, Hany Abdulsamad, Boris Belousov, Carlo D'Eramo, Jan Peters, Joni Pajarinen
Across machine learning, the use of curricula has shown strong empirical potential to improve learning from data by avoiding local optima of training objectives.
no code implementations • 1 Jul 2020 • Tuan Dam, Carlo D'Eramo, Jan Peters, Joni Pajarinen
Monte-Carlo planning and Reinforcement Learning (RL) are essential to sequential decision making.
1 code implementation • NeurIPS 2020 • Pascal Klink, Carlo D'Eramo, Jan Peters, Joni Pajarinen
Curriculum reinforcement learning (CRL) improves the learning speed and stability of an agent by exposing it to a tailored series of tasks throughout learning.
no code implementations • 20 Mar 2020 • Andrea Cini, Carlo D'Eramo, Jan Peters, Cesare Alippi
In this regard, Weighted Q-Learning (WQL) effectively reduces bias and shows remarkable results in stochastic environments.
2 code implementations • 4 Jan 2020 • Carlo D'Eramo, Davide Tateo, Andrea Bonarini, Marcello Restelli, Jan Peters
MushroomRL is an open-source Python library developed to simplify the process of implementing and running Reinforcement Learning (RL) experiments.
1 code implementation • 1 Jan 2020 • Simone Parisi, Davide Tateo, Maximilian Hensel, Carlo D'Eramo, Jan Peters, Joni Pajarinen
Empirical results on classic and novel benchmarks show that the proposed approach outperforms existing methods in environments with sparse rewards, especially in the presence of rewards that create suboptimal modes of the objective function.
no code implementations • 1 Nov 2019 • Tuan Dam, Pascal Klink, Carlo D'Eramo, Jan Peters, Joni Pajarinen
Finally, we empirically demonstrate the effectiveness of our method in well-known MDP and POMDP benchmarks, showing significant improvement in performance and convergence speed w. r. t.