Juan Hernandez

Juan Hernandez, Diego Fernández, Manuel Cifuentes et al.

The Never Give Up (NGU) algorithm has proven effective in reinforcement learning tasks with sparse rewards by combining episodic novelty and intrinsic motivation. In this work, we extend NGU to multi-agent environments and evaluate its performance in the simple_tag environment from the PettingZoo suite. Compared to a multi-agent DQN baseline, NGU achieves moderately higher returns and more stable learning dynamics. We investigate three design choices: (1) shared replay buffer versus individual replay buffers, (2) sharing episodic novelty among agents using different k thresholds, and (3) using heterogeneous values of the beta parameter. Our results show that NGU with a shared replay buffer yields the best performance and stability, highlighting that the gains come from combining NGU intrinsic exploration with experience sharing. Novelty sharing performs comparably when k = 1 but degrades learning for larger values. Finally, heterogeneous beta values do not improve over a small common value. These findings suggest that NGU can be effectively applied in multi-agent settings when experiences are shared and intrinsic exploration signals are carefully tuned.

Javier Rojo, Juan Hernandez, Juan Manuel Murillo et al.

Today's world is a globalized and connected one, where people are increasingly moving around and interacting with a greater number of services and devices of all kinds, including those that allow them to monitor their health. However, each company, institution or health system usually store its patients' data in an isolated way. Although this approach can have some benefits related with privacy, security, etc., it also implies that each one of them generates different, incomplete and possibly contradictory views of a patient's health data, losing part of the value that this information could bring to the patient. That is the reason why researchers from all over the world are determined to replace the current institution-centered health systems with new patient-centered ones. In these new systems, all the health information of a patient is integrated into a unique global vision. However, some questions are still unanswered. Specifically, who should store and maintain the information of a given patient and how should this information be made available for other systems. To address this situation, this work proposes a new solution towards making the Personal Health Trajectory of patients available for both, the patients themselves and health institutions. By using the concept of blockchains' federation and web services access to the global vision of a person health can be granted to existing and new solutions. To demonstrate the viability of the proposal, an implementation is provided alongside the obtained results in a potential scenario.