Named entity recognition (NER) is a key component of many scientific literature mining tasks, such as information retrieval, extraction, and question answering; however, modern approaches require large amounts labeled training data in order to be effective. This severely limits the effectiveness NER models applications where expert annotations are difficult expensive obtain. In this work, we explore transfer learning semi-supervised self-training improve performance biomedical settings with...

Today's cities generate tremendous amounts of data, thanks to a boom in affordable smart devices and sensors. The resulting big data creates opportunities develop diverse sets context-aware services systems, ensuring city are optimized the dynamic environment. Critical resources these will be more rapidly deployed regions need, those predicted have an imminent or prospective need. For example, crime analytics may used optimize distribution police, medical, emergency services. However, as...

Intelligent Heating, Ventilation, and Air Conditioning (HVAC) control using deep reinforcement learning (DRL) has recently gained a lot of attention due to its ability optimally the complex behavior HVAC system. However, more exploration is needed on understanding adaptability challenges that DRL agent could face during deployment phase. Using online for such applications not realistic long period likely poor comfort process. Alternatively, can be pre-trained building model prior deployment....

Existing approaches for allocating resources on edge environments are inefficient and lack the support of heterogeneous devices, which in turn fail to optimize dependency cloud infrastructures or datacenters. To this extent, we propose paper OpERA, a multi-layered edge-based resource allocation optimization framework that supports seamless execution offloadable tasks across edge, fog, computing layers architectures. By capturing task requirements, OpERA is capable identifying suitable within...

Deep reinforcement learning (DRL) approaches have been used in various application areas to improve efficiency, optimization, or automation. However, very little is known about how the DRL algorithms make decisions and what features affect their performance. Using a case study of based Heating, Ventilation Air Conditioning (HVAC) optimization methodology, we demonstrate can address these challenges by applying interpretability tools systematically exploring model inputs for better...

Recent advances in wireless sensor networks (WSN)technologies are enabling patient-centered, result oriented health care services, the most cost-effective manner. Furthermore, technological a greater shift from institutional services to community-based services. The model of collaborative health system (COHESY) presented this paper offers 24 hour monitoring the condition patients and possibility sending an emergency call for sudden deterioration his medical condition. In addition,...

In this study, we present a framework based on prediction model that facilitates user access to number of services in smart living environment. Users must be able all available continuously equipped with mobile devices or objects without being impacted by technical constraints such as performance memory issues, regardless their physical location and mobility. To achieve goal, propose the use cloudlet-based architecture serves distributed cloud resources specific ranges influence realtime...

The Artificial Intelligence (AI) development described herein uses model-free Deep Reinforcement Learning (DRL) to minimize energy cost during residential heating, ventilation, and air conditioning (HVAC) operation. Building cooling loads HVAC operation are difficult accurately model due complexity, lack of measurements data, specific performance, so online machine learning is used allow for real-time readjustment in performance. Energy costs the multi-zone unit shown this work minimized by...

Internet of Things (IoT) is becoming more pervasive in many installations, including homes, manufacturing plants, and industrial facilities all kinds. The data that IoT produces a reflection usual behavior such as daily routines scheduled tasks, but also from unexpected due to unintentional or undesirable abnormalities. Here, we focus on achieving coordinated intelligence about normal abnormal phenomena multiple sensors are geographically co-located close proximity, monitoring controlling...

Recently, deep reinforcement learning (DRL) based intelligent control of Heating, Ventilation, and Air Conditioning (HVAC) has gained a lot attention due to DRL's ability optimally HVAC for minimizing operational cost while maintaining resident's comfort. The success such DRL-based techniques largely depends on the articulation problem in terms states, actions, reward function. Inclusion electricity pricing information formulation can play an important role saving operation. However, less...

This paper gives an understanding of what possibilities Wireless Body Area Network (WBAN) have when using shortrange wireless communications protocols. There advantages are used to facilitate versatility in the movements health care patients. The investigates feasibility ZigBee protocol, givean analysis methods for collection received data from multi sensor environment, and mechanisms privacy protection by encryption techniques. characteristics applied on previously developed collaborative...

The cybersecurity auditing for Operation Technology is critical and has been largely missing from the research, especially in energy sector. In this paper, we present a novel "cybersecurity vetting" approach (CYVET) to problem of verification validation complex cyber-physical installations underlying modern grid systems.

As the number of online services has increased, amount sensitive data being recorded is rising. Simultaneously, decision-making process improved by using vast amounts data, where machine learning transformed entire industries. This paper addresses development optimal private deep neural networks and discusses challenges associated with this task. We focus on differential privacy implementations finding balance between accuracy privacy, benefits limitations existing libraries, applying models...

The population is increasingly becoming tractable as more and people carry handheld devices part of their everyday activities. Recent studies have shown that devices' generated traffic share now than 50% total global online traffic. This has created an unprecedented opportunity for modeling human mobility behavior. For example, aggregate check-ins dwell time can reveal building level occupancies. However, there are clear limits to accurate (e.g. reproducible, repeatable, realistic), unless...

Graph Neural Networks (GNNs) have gained significant attention owing to their ability handle graph-structured data and the improvement in practical applications. However, many of these models prioritize high utility performance, such as accuracy, with a lack privacy consideration, which is major concern modern society where attacks are rampant. To address this issue, researchers started develop privacy-preserving GNNs. Despite progress, there comprehensive overview techniques for preserving...