Due to the recent progress in Deep Neural Networks, Reinforcement Learning (RL) has become one of most important and useful technology. It is a learning method where software agent interacts with an unknown environment, selects actions, progressively discovers environment dynamics. RL been effectively applied many areas real life. This article intends provide in-depth introduction Markov Decision Process, its algorithms. Moreover, we present literature review application variety fields,...

Modern Medical Information Systems very often comprise Devices and governed by regulations which require stringent Risk Management activities to be implemented minimize the occurrence of safety risks. Currently, reference standard adopted manufacturers for is ISO 14971, which, however, was devised traditional (mostly hardware) does not either take into account peculiarities modern Systems, or define a formal methodology conduct Assessment. Moreover, approaches currently typically aims at...

We introduce fractal social organizations—a novel class of socio‐technical complex systems characterized by a distributed, bio‐inspired, hierarchical architecture. Based on same building block that is recursively applied at different layers, said provide homogeneous way to model collective behaviors complexity and scale. Key concepts principles are enunciated means case study simple formalism. As preliminary evidence the adequacy assumptions underlying our here, we define an algebraic for...

Parkinson’s disease (PD) is a cognitive degenerative disorder of the central nervous system that mainly affects motor system. The earliest symptoms evidence general deficit coordination and an unsteady gait. Current approaches for evaluation assessment gait disturbances in PD have proved to be expensive, inconvenient ineffective detection anomalous walking patterns. In this paper, we address these issues by defining deep time series-based approach patterns dynamics elderly people analyzing...

Time Series Forecasting (TSF) is an important application across many fields. There a debate about whether Transformers, despite being good at understanding long sequences, struggle with preserving temporal relationships in time series data. Recent research suggests that simpler linear models might outperform or least provide competitive performance compared to complex Transformer-based for TSF tasks. In this paper, we propose novel data-efficient architecture, GLinear, multivariate exploits...

The design of ambient intelligence applications in critical systems requires rigorous software-engineering-oriented approaches. Drawing on practical experience, the authors propose a set formal tools and specification process for AmI activities artifacts.

Wireless and pervasive healthcare applications typically present critical requirements from the point of view functional correctness, reliability, availability, security, safety. In contrast to case classic safety applications, behavior wireless is affected by movements location users resources. This article presents a methodology formally express in order achieve higher degree dependability. particular, it will be shown how possible formalize constrict mobility characteristics combining,...

Ambient Intelligence technologies have not yet been widely adopted in safety critical scenarios. This principally has due to fact that acceptable degrees of dependability reached for the applications rely on such technologies. However, new application domains, like Assisted Living and Smart Hospitals, which are currently emerging, increasing need methodologies tools can improve reliability final systems. paper presents a middleware architecture provides developer with services runtime...

In recent decades the aging of population has led to a change in health held with special attention issue home care and e-health. The aim is provide different types services patient's rather than hospital so improving quality life patients by allowing them stay their own environment. healthcare sector undergoing profound transformation thanks possibilities offered Internet Things new technologies, mobile wearable. model oriented overall patient, stimulated implemented through strong pro...

This article presents an Artificial Intelligence (AI)-based infrastructure to reduce medication errors while following a treatment plan at home. The system, in particular, assists patients who have some cognitive disability. AI-based system first learns the skills of patient using Actor–Critic method. After assessing patients’ disabilities, adopts appropriate method for monitoring process. Available methods process are Deep Learning (DL)-based classifier, Optical Character Recognition, and...

Dynamic Treatment Regimes (DTRs) are sets of sequential decision rules that can be adapted over time to treat patients with a specific pathology.DTR consists alternative treatment paths and any these treatments depending on the patient's characteristics.Reinforcement Learning (RL) Imitation (IL) approaches have been deployed for obtaining optimal patient but, rely only positive trajectories (i.e., concluded responses patient).In contrast, negative samples non-responding treatments)...