Ioannis D. Zaharakis

Machine learning: a review of classification and combining techniques

Supervised classification is one of the tasks most frequently carried out by so-called Intelligent Systems. Thus, a large number of techniques have been developed based on Artificial Intelligence (Logic-based techniques, Perceptron-based techniques) and Statistics (Bayesian Networks, Instance-based techniques). The goal of supervised learning is to build a concise model of the distribution of class labels in terms of predictor features. The resulting classifier is then used to assign class labels to the testing instances where the values of the predictor features are known, but the value of the class label is unknown. This paper describes various classification algorithms and the recent attempt for improving classification accuracy—ensembles of classifiers.

Bagged Averaging of Regression Models

Linear regression and regression tree models are among the most known regression models used in the machine learning community and recently many researchers have examined their sufficiency in ensembles. Although many methods of ensemble design have been proposed, there is as yet no obvious picture of which method is best. One notable successful adoption of ensemble learning is the distributed scenario. In this work, we propose an efficient distributed method that uses different subsets of the same training set with the parallel usage of an averaging methodology that combines linear regression and regression tree models. We performed a comparison of the presented ensemble with other ensembles that use either the linear regression or the regression trees as base learner and the performance of the proposed method was better in most cases.

Cryptography and Security in Internet of Things (IoTs): Models, Schemes, and Implementations

Internet of Things (IoT) has attracted the interest of the research community, at a great manner the last years. It is one of the most hot topics, for current and future research areas, involved by both academia and industry sectors. Security and privacy issues for IoT are proven targets of great importance. This tutorial brings together models, schemes and implementation aspects, regarding the alternative technologies and devices of IoT. It focusses on the knowledge of the implementation efficiency, basically in hardware aspects, of communications confidentiality, user authentication, data integrity and services availability. Attacks and modern threats, as well as to countermeasures against them, are considered. Different IoT infrastructures are presented and alternative trust models are illustrated. Policy issues are also discussed.

A Study of Bio-inspired Communication Scheme in Swarm Robotics

In this study two bio-inspired communication schemes (stigmergy and hormone based communication) are evaluated according to the contribution of each to the self organisation of a multi agent robotic system. The evaluation is carried on by the simulation of a test case scenario and it is based on measurable criteria that only depend on triggered agent actions. The results present the different situations where each of the communication schemes performs better in terms of task accomplishment.

Modeling spiking neural networks

A notation for the functional specification of a wide range of neural networks consisting of temporal or non-temporal neurons, is proposed. The notation is primarily a mathematical framework, but it can also be illustrated graphically and can be extended into a language in order to be automated. Its basic building blocks are processing entities, finer grained than neurons, connected by instant links, and as such they form sets of interacting entities resulting in bigger and more sophisticated structures. The hierarchical nature of the notation supports both top-down and bottom-up specification approaches. The use of the notation is evaluated by a detailed example of an integrated tangible agent consisting of sensors, a computational part, and actuators. A process from specification to both software and hardware implementation is proposed.

Social intelligence as the means for achieving emergent interactive behaviour in ubiquitous computing environments

This work introduces a framework for modelling the main actors (human, artefacts and services) in a symbiotic Ambient Intelligence environment. It, also, proposes an architectural scheme that associates the social behaviour, which is not an inherent characteristic of the participants, during interaction with the functional behaviour of the participants of a Ubiquitous Computing application. The overall approach is demonstrated by a specific example of application which illustrates its concepts through a more technical point of view.

An instructional design process for creating a U-learning ecology

The ubiquitous computing (UbiComp) is considered as an extension of the computational capabilities of the physical environment, allowing the computational structure to be present everywhere in the form of small, robust, networked processing devices distributed at all scales through everyday life and generally turned to distinctly common place ends. There are various research challenges regarding the design and use of instructional design tools in complex learning contexts such as Ubiquitous Computing, Mobile learning (m-learning) and Internet of Things (IoT), the technologies defined as UMI technologies. This paper presents the rationale, important issues and methodology constructed in the context of UbiComp so as to initially define an instructional design process for building a U — Learning Ecology for multidisciplinary education. We provide a consistent framework and structural view of integrating instructional design principles in UbiComp learning: we discuss our ideas on the design of a U-learning ecology by the gradual building of a robust design process and we provide an overview of our ongoing work on design/analysis tools supporting early stage prototyping for using UMI technologies.

Security & Trusted Devices in the Context of Internet of Things (IoT)

This work targets to the technologies of the Internet of Things (IoT), regarding to security and trusted devices. It provides to the readers a comprehensive understanding of both security and privacy aspects. Modern systems and networks are quoted, in order to cover any questions arising from the theoretical approach. Hardware integration devices are also presented, for flexible implementations for the presented IoT technologies. Emphasis is given to IoT embedded hardware platforms like Udoo, which fully support IoT implementations. Last but not least, data and information preservation are analyzed, so as not to get lost or misused.

An Implementation of a Symbiotic Ecology with Self*-Properties in an eHealth Scenario

This paper presents an implementation of a symbiotic ecology with self*-properties in an eHealth scenario. In addition, presents the architecture of a swarm which consists of a small number of artefacts that operate in the symbiotic ecology. The aim of this work is to study the emergence of self*-properties and particularly the influence of the environment on the timely achievement of the objectives of the swarm. A series of experiments are presented, focusing at the observation of self*-properties and the response times of the swarm. All experiments are performed with the use of a simulator.

A hybrid expert system as an embedded module in tutoring systems

Abstract Every Intelligent Tutoring System (ITS), in order to justify the term ‘intelligent’, has to contain the specific knowledge of the taught subject as well as to embed several modules with those mechanisms that are able to handle the knowledge in an intelligent way (e.g., using rules of thumb, tricks etc.) and, in general, behave in a way that would be considered intelligent if observed by a human. Such modules are apparently necessary in ITS generators too. In this paper, the Hybrid Methodology Tutor (HMeT) is presented. HMeT is a hybrid expert system that is embedded in an ITS generator and supports the description of the structure and dynamics of the procedural knowledge, during authoring; it is also responsible for its presentation during tutoring. HMeT was designed and implemented to replace an earlier version (MeT ‐ Methodology Tutor), in order to provide for higher inferential capabilities, better user interface and more user friendliness in authoring as well as in tutoring. HMeT functions li...