Filetype
MushroomRL is an open-source Python library developed to simplify the process of im- plementing and running Reinforcement Learning (RL) experiments. Compared to other available libraries, MushroomRL has been created with the purpose of providing a com- prehensive and exible framework to minimize the effort in implementing and testing novel RL methodologies. The architecture of MushroomRL is built
Highly dynamic robotic tasks require high-speed and reactive robots. These tasks are particularly challenging due to the physical constraints, hardware limitations, and the high uncertainty of dynamics and sensor measures. To face these issues, it's crucial to design robotics agents that generate precise and fast trajectories and react immediately to environmental changes. Air hockey is an example
Reinforcement learning in robotics is extremely challenging due to many practical issues, including safety, mechanical constraints, and wear and tear. Typically, these issues are not considered in the machine learning literature. One crucial problem in applying reinforcement learning in the real world is Safe Exploration, which requires physical and safety constraints satisfaction throughout the l
This thesis investigates the role of trust in the context of decentralised finance (DeFi). Representing a new paradigm shift in finance, DeFi emerge from the technological developments that we have witnessing in the last decades, particularly decentralised ledger technologies (DLTs) of which Blockchain is perhaps the most popular. DeFi is changing the way that firms raise financial capital and a c
This paper examines how the concept ‘nature-positive’ as a means to enhance biodiversity is defined and used, viewed through the lens of power. Building on a three-dimensional conceptualization of power, we elaborate on i) how ‘nature-positive’ enters and remains on business and policy agendas, ii) different interpretations from both ecological and business perspectives, and iii) the governance of
Transferring patients from the ward of their specialty or consultant is described as boarding. 1 Boarding patients is becoming increasingly prevalent due to greater pressure on hospital capacity. This practice compromises patient safety through delayed investigations, prolonged hospital stays, and increased risk of hospital-acquired infections. 1 2 We evaluated how regularly boarded patients were
Reinforcement learning methods for robotics are increasingly successful due to the constant development of better policy gradient techniques. A precise (low variance) and accurate (low bias) gradient estimator is crucial to face increasingly complex tasks. Traditional policy gradient algorithms use the likelihood-ratio trick, which is known to produce unbiased but high variance estimates. More mod
We report an unusual case of a patient presenting with Cushing’s syndrome caused by a phaeochromocytoma secreting adrenocorticotropic hormone (ACTH). The patient had a history of treatment-resistant hypertension, secondary amenorrhoea and tendency towards hypokalaemia. She had multiple signs of Cushing’s syndrome, such as swelling, bruising, abdominal striae and proximal myopathy. Hypokalaemia is
Global environmental challenges, coupled with growing material resource consumption, are exacerbated by an increasingly hostile geopolitical environment and the securitisation of politics. The circular economy (CE) proposes a solution to address these challenges by transforming production and consumption systems, which has sparked significant interest in the concept among scholars, policymakers, a
This study protocol outlines the test evaluation of the design of a comprehensive Nordic hospital at home (HaH) digital education program entitled “the Nordic Digital Health and Education” program (NorDigHE). The NorDigHE program aims to prepare healthcare professionals for healthcare provision in a HaH context. The program, developed by a Nordic consortium, spans five modules over 3.5 weeks of fu
We study the benefit of sharing representations among tasks to enable the effective use of deep neural networks in Multi-Task Reinforcement Learning. We leverage the assumption that learning from different tasks, sharing common properties, is helpful to generalize the knowledge of them resulting in a more effective feature extraction compared to learning a single task. Intuitively, the resulting s
We introduce ImitationFlow, a novel Deep generative model that allows learning complex globally stable, stochastic, nonlinear dynamics. Our approach extends the Normalizing Flows framework to learn stable Stochastic Differential Equations. We prove the Lyapunov stability for a class of Stochastic Differential Equations and we propose a learning algorithm to learn them from a set of demonstrated tr
There is an increasing interest in Reinforcement Learning to solve new and more challenging problems, as those emerging in robotics and unmanned autonomous vehicles. To face these complex systems, a hierarchical and multi-scale representation is crucial. This has brought the interest on Hierarchical Deep Reinforcement learning systems. Despite their successful application, Deep Reinforcement Learn
ObjectiveAs part of our research on Swedish school principals, we examined the occurrence and co-occurrence of conflicts and five negative acts (i.e. harassment, sexual harassment, threats of violence or physical harm, physical violence, bullying), their attributed sources, and their associations with self-rated health, workability, and general life satisfaction among 2670 principals.ResultsDuring
Recognizing individual variability is essential for developing targeted, personalized medical interventions. Vocal fatigue is a prevalent symptom and complaint among occupational voice users, but its identification has yielded mixed results. Vocal fatigue is a complex issue with heterogeneous biophysiological responses to vocal demands among individuals. This research aims to classify individuals
Study region: The Paranapanema River Basin, located in southeastern Brazil, is characterized by a cascade of large hydropower plants regulated by a nationally coordinated dispatch system. This basin is a representative case of reservoir-regulated rivers in the country, where multiple dams interact to supply electricity while reshaping natural flow regimes. Study focus: This study examines hydropow
Against the backdrop of the ecological crisis, food production systems will need to transform to ensure future food security. There are contrasting visions of what this transformation should look like. This paper scrutinizes these competing imaginaries of future agriculture and associated styles of anticipation in Tanzania. We find that both government visions and global crop models anticipate fut
Purpose: Children with delayed reconstruction of esophageal atresia (DREA) have a high risk of morbidity. The aim of this study was to determine the prevalence of feeding difficulties in children with DREA, differences to children with primary anastomosis (PA) and identify risk factors for feeding difficulties in children with DREA. Methods: Parents of 30 children with DREA, born 2001–2018, were r
