Abstract
When compared with their single-agent counterpart, multi-agent systems have an additional set of challenges for reinforcement learning algorithms, including increased complexity, non-stationary environments, credit assignment, partial observability, and achieving coordination. Deep reinforcement learning has been shown to achieve successful policies through implicit coordination, but does not handle partial-observability. This paper describes a deep reinforcement learning algorithm, based on multi-agent actor-critic, that simultaneously learns action policies for each agent, and communication protocols that compensate for partial-observability and help enforce coordination. We also research the effects of noisy communication, where messages can be late, lost, noisy, or jumbled, and how that affects the learned policies. We show how agents are able to learn both high-level policies and complex communication protocols for several different partially-observable environments. We also show how our proposal outperforms other state-of-the-art algorithms that don't take advantage of communication, even with noisy communication channels.

Abstract
In a deeply interconnected world of people and goods, infectious diseases constitute a serious threat. An active vigilance is required. The collection of adequate data is vital and coordinated by surveillance systems. It is widely-acknowledged that every case-reporting system has some degree of under-reporting and reporting delay in particular in HIV-AIDS Portuguese Surveillance System. To better understand the processes generating the reporting delays, which is an administrative process, it was used a flexible continuous time fully parametric survival analysis approach. It was taken into consideration the hierarchical administrative and organizational structure of the system as well as the relevant changes in the procedures throughout the time. The best multilevel structure to represent reporting delays in continuous time is the model where the individuals are nested into Reporting Entities (20.24% of the variance) which are nested into Type of services (8% of the variance) with the log-normal distribution. © 2019, Springer Nature Switzerland AG.

Abstract
Surveillance has been defined as the continual scrutiny of all aspects in emerging and the spread of a disease that is pertinent to effective control, involving a systematic collection, analysis, interpretation, and dissemination of health data. Given their fragmentation several problems inherent to data must be recognized. This paper aims to provide an overview of European Public Health Surveillance Systems emphasizing their structure and main challenges. The HIV-AIDS surveillance is overview as a particular case. The most common issues are unrepresentativeness, changes in the implementation through time, inconsistent use of case definitions, miss diagnoses, miss or fail to report a case, reporting delay, and errors during completion of the form or data entry. The HIV - AIDS surveillance is one of the most complex mainly due to the special epidemiology of the disease surrounding the transmission modes and the lack of treatment and all the socio-ecological framework involved. © Springer Nature Switzerland AG 2019.

Abstract
A solution for an optimal robotic soccer strategy is yet to be found. Multiple agents interacting in an environment with continuous state and action spaces is a recipe for machine learning lethargy. Fortunately, due to the increasing hardware performance and new algorithms, the area of reinforcement learning is growing considerably. However, simulating the environment for strategy purposes is not following this trend. Several simulators are available, including those used in major soccer competitions (e.g. RoboCup). However, no option combines a good repository of teams with a simple command set that abstracts low-level actions. To clarify this problem, we surveyed the most promising simulators and proposed an extension for the well-established Soccer Server. The objective was to simplify the process of learning strategy-related behaviors through automated optimization algorithms. The results have shown a clear advantage in using the extension to improve the agent's performance. This work contributes to the development of future strategies related with RoboCup or other soccer competitions. Despite the good results, there is space for improvement in computational efficiency and behavior diversity.

Abstract
The Weighted Policy Learner (WPL) algorithm has been shown to converge to Nash Equilibria (NE) in several challenging environments with minimum knowledge. However, WPL has trouble converging to deterministic strategies, since the policy update rate approaches zero. We propose a new update rule that bounds this update rate such that, in pure NE games, the algorithm’s speed is not slowed down, while its behavior in stochastic NE games remains unchanged. We demonstrate our proposal’s behavior in several common game-theoretic environments (with stochastic and deterministic equilibrium policies), in complex maze-related games (where some actions dominate others in most states), against the original WPL as well as other state of the art algorithms. We draw conclusions over the benefits of our solution and its advantages. © Springer Nature Switzerland AG 2019.

Abstract
Cancer is a serious disease that causes significant disability and suffering, so naturally Health Related Quality of Life (HRQoL) is a major concern of patients, families and clinicians. This paper intends to relate biometric indices, in terms of HRV metrics, with self-perceived HRQoL from patients with lymphoma. Patients (N = 12) answered FACT questionnaire and used a smartband that collected biometrical data in real-time along the chemotherapy treatment. Our results revealed that Physical Well-Being, Total, Lymphoma subscale and FACT-Lym Trial Outcome domains seem to have a similar pattern that HRV metrics across the treatment cycles. In specific, the FACT domains and the HRV metrics have the lowest average levels on the first cycle and seem to increase along the following cycles (3 rd and 6 th cycles). This approach of continuous assessment of HRQoL will enable a better accuracy and more supported clinical decision. © Springer Nature Switzerland AG 2019.