Showing papers by "Lakhmi C. Jain published in 2013"

A Deterministic Improved Q-Learning for Path Planning of a Mobile Robot

Abstract: This paper provides a new deterministic Q-learning with a presumed knowledge about the distance from the current state to both the next state and the goal. This knowledge is efficiently used to update the entries in the Q-table once only by utilizing four derived properties of the Q-learning, instead of repeatedly updating them like the classical Q-learning. Naturally, the proposed algorithm has an insignificantly small time complexity in comparison to its classical counterpart. Furthermore, the proposed algorithm stores the Q-value for the best possible action at a state and thus saves significant storage. Experiments undertaken on simulated maze and real platforms confirm that the Q-table obtained by the proposed Q-learning when used for the path-planning application of mobile robots outperforms both the classical and the extended Q-learning with respect to three metrics: traversal time, number of states traversed, and 90° turns required. The reduction in 90° turnings minimizes the energy consumption and thus has importance in the robotics literature.

Realization of an Adaptive Memetic Algorithm Using Differential Evolution and Q-Learning: A Case Study in Multirobot Path Planning

Abstract: Memetic algorithms (MAs) are population-based meta-heuristic search algorithms that combine the composite benefits of natural and cultural evolutions. An adaptive MA (AMA) incorporates an adaptive selection of memes (units of cultural transmission) from a meme pool to improve the cultural characteristics of the individual member of a population-based search algorithm. This paper presents a novel approach to design an AMA by utilizing the composite benefits of differential evolution (DE) for global search and Q-learning for local refinement. Four variants of DE, including the currently best self-adaptive DE algorithm, have been used here to study the relative performance of the proposed AMA with respect to runtime, cost function evaluation, and accuracy (offset in cost function from the theoretical optimum after termination of the algorithm). Computer simulations performed on a well-known set of 25 benchmark functions reveal that incorporation of Q-learning in one popular and one outstanding variants of DE makes the corresponding algorithm more efficient in both runtime and accuracy. The performance of the proposed AMA has been studied on a real-time multirobot path-planning problem. Experimental results obtained for both simulation and real frameworks indicate that the proposed algorithm-based path-planning scheme outperforms the real-coded genetic algorithm, particle swarm optimization, and DE, particularly its currently best version with respect two standard metrics defined in the literature.

Local Binary Patterns: New Variants and Applications

Abstract: This book introduces Local Binary Patterns (LBP), arguably one of the most powerful texture descriptors, and LBP variants. This volume provides the latest reviews of the literature and a presentation of some of the best LBP variants by researchers at the forefront of textual analysis research and research on LBP descriptors and variants. The value of LBP variants is illustrated with reported experiments using many databases representing a diversity of computer vision applications in medicine, biometrics, and other areas. There is also a chapter that provides an excellent theoretical foundation for texture analysis and LBP in particular. A special section focuses on LBP and LBP variants in the area of face recognition, including thermal face recognition. This book will be of value to anyone already in the field as well as to those interested in learning more about this powerful family of texture descriptors.

Monitoring pilot “Situation Awareness”

Abstract: Pilot error remains the major cause of aircraft accidents. The lack of Situation Awareness SA, even amongst experienced pilots, is one of the primary reasons for pilot error. It is important to ensure that pilots are able to maintain a high level of SA before they act as members of a flight crew. A pilot's SA can be assessed by monitoring pilot behavior using observations of pilot eye movement. In this paper we discuss a rule-based approach to assess novice pilots' SA against a baseline. A recent experiment confirmed there is a measurable difference between the eye movement of an experienced pilot and the eye movement of several novice pilots. Initially, an expert pilot's eye movement was recorded using an eye tracker device in order to set the baseline. A Gaze Analyzer was used to derive relevant information from eye movement data. After consulting a Subject Matter Expert SME, a rule-based system was created to monitor the pilot's behavior. The data was analyzed to identify behavioral patterns in order to associate the pilots' level of SA with their mode of attention distribution fixation. Novice pilot's eye movement data was compared with an expert pilot's eye movement data using an inference engine.

Multiagent Systems and Applications: Volume 2 Development Using the GORITE BDI Framework

Abstract: Since its conception almost 30 years ago, the BDI (Belief Desire Intention) model of agency has become established, along with Soar, as the approach of choice for practitioners in the development of knowledge intensive agent applications. However, in developing BDI agent applications for over 15 years, the authors of this book have observed a disconnect between what the BDI model provides and what is actually required of an agent model in order to build practical systems. The GORITE BDI framework was developed to address this gap and this book is written for students, researchers and practitioners who wish to gain a practical understanding of how GORITE is used to develop BDI agent applications. In this regard, a feature of the book is the use of complete, annotated examples. As GORITE is a Java framework, a familiarity with Java (or a similar language) is assumed, but no prior knowledge of the BDI model is required.

Sustainability in Energy and Buildings : Proceedings of the 4th International Conference in Sustainability in Energy and Buildings (SEB´12)

Abstract: Sustainability in Energy and Buildings : Proceedings of the 4th International Conference in Sustainability in Energy and Buildings (SEB´12)

Issues and Trends

Abstract: From the content: Advanced Query Processing: An Introduction.- On Skyline Queries and how to Choose from Pareto Sets.- Processing Framework for Ranking and Skyline Queries.- Preference-Based Query Personalization.- Approximate Queries for Spatial Data.- Approximate XML Query Processing.- Progressive and Approximate Join Algorithms on Data Streams.

Knowledge Engineering, Machine Learning and Lattice Computing with Applications: 16th International Conference, KES 2012, San Sebastian, Spain, ... / Lecture Notes in Artificial Intelligence)

Abstract: This book constitutes the refereed proceedings of the 16th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems, KES 2012, held in San Sebastian, Spain, in September 2012. The 20 revised full papers presented were carefully reviewed and selected from 130 submissions. The papers are organized in topical sections on bioinspired and machine learning methods, machine learning applications, semantics and ontology based techniques, and lattice computing and games.

New Trends in Web User Behaviour Analysis

Abstract: The analysis of human behaviour has been conducted within diverse disciplines, such as psychology, sociology, economics, linguistics, marketing and computer science. Hence, a broad theoretical framework is available, with a high potential for application into other areas, in particular to the analysis of web user browsing behaviour. The above mentioned disciplines use surveys and experimental sampling for testing and calibrating their theoretical models. With respect to web user browsing behaviour, the major source of data is the web logs, which store every visitor’s action on a web site. Such files could contain millions of registers, depending on the web site traffic, and represents a major data source about human behaviour. This chapter surveys the new trends in analysing web user behaviour and revises some novel approaches, such as those based on the neurophysiological theory of decision making, for describing what web users are looking for in a web site.

An Introduction to Quality Issues in the Management of Web Information

Abstract: Since a long time information Quality Assessment is a crucial issue in organizations, where it relates to the ability of the organization to adequately fulfil the needs and expectations of its customers and users (Batini et al. 2008). In this context the evaluation of data quality has been the main concern (Batini and Scannapieco 2006).

Multimedia Services in Intelligent Environments: Recommendation Services

Abstract: Multimedia services are now commonly used in various activities in the daily lives of humans. Related application areas include services that allow access to large depositories of information, digital libraries, e-learning and e-education, e-government and e-governance, e-commerce and e-auctions, e-entertainment, e-health and e-medicine, and e-legal services, as well as their mobile counterparts (i.e., m-services). Despite the tremendous growth of multimedia services over the recent years, there is an increasing demand for their further development. This demand is driven by the ever-increasing desire of society for easy accessibility to information in friendly, personalized and adaptive environments. In this book at hand, we examine recent Recommendation Services. Recommendation services appear in the mobile environment, medicine/biology, tourism, education, and so on. The book includes ten chapters, which present various recently developed recommendation services.This research book is directed to professors, researchers, application engineers and students of all disciplines who are interested in learning more about recommendation services, advancing the corresponding state of the art and developing innovative recommendation services.

Extending the Contraposition Property of Propositional Logic for Fuzzy Abduction

Abstract: Abduction deals with assumption-based reasoning to explain an observation. In the context of fuzzy reasoning, abduction attempts to determine the membership function of the fuzzy propositions present in the antecedent of a rule when the membership functions for the propositions in the consequent of the rule are given. Currently available models of fuzzy abduction are capable of inferring the membership function of the antecedent clause accurately when the antecedent includes single fuzzy proposition. However, when the antecedent clause of a rule contains multiple fuzzy propositions, these models fail to determine the independent membership function of the individual propositions present in the antecedent. This paper presents a new formulation to handle the above problem by fuzzy extension of the well-known contraposition property of propositional logic. Several interesting properties due to the fuzzy extension of the classical contraposition have been derived. An algorithm for automated abduction using the extended contraposition property has been developed to demonstrate the principle of abduction with rules containing one or more fuzzy propositions in the antecedent/consequent. The time complexity of the proposed fuzzy abduction for a sequence of n-chained rules, where each rule has m fuzzy propositions, is O (mn), considering a uniform cost for composition operation and t-norm computation of the antecedent.

Multimedia Services in Intelligent Environments: Advances in Recommender Systems

Abstract: Multimedia services are now commonly used in various activities in the daily lives of humans. Related application areas include services that allow access to large depositories of information, digital libraries, e-learning and e-education, e-government and e-governance, e-commerce and e-auctions, e-entertainment, e-health and e-medicine, and e-legal services, as well as their mobile counterparts (i.e., m-services). Despite the tremendous growth of multimedia services over the recent years, there is an increasing demand for their further development. This demand is driven by the ever-increasing desire of society for easy accessibility to information in friendly, personalized and adaptive environments. In this book at hand, we examine recent Advances in Recommender Systems. Recommender systems are crucial in multimedia services, as they aim at protecting the service users from information overload. The book includes nine chapters, which present various recent research results in recommender systems. This research book is directed to professors, researchers, application engineers and students of all disciplines who are interested in learning more about recommender systems, advancing the corresponding state of the art and developing recommender systems for specific applications.

Advanced Query Processing: An Introduction

Abstract: Traditional query processing techniques have played a major role in the success of relational Database Management Systems over the last decade. However, they do not obviously extend to much more challenging, unorganized and unpredictable data providers, typical of emerging data intensive applications and novel processing environments. For them, advanced query processing and data integration approaches have been proposed with the aim of still guaranteeing an effective and efficient data access in such more complex data management scenarios. The aim of this chapter is to present the main issues and trends arising in advanced query processing and to relate them to the various parts of this book. For each part, a brief description of the background concepts and of the presented contributions is also provided.

Emerging Paradigms in Machine Learning: An Introduction

Abstract: This chapter provides a broad overview of machine learning (ML) paradigms both emerging as well as well-established ones These paradigms include: Bayesian Learning, Decision Trees, Granular Computing, Fuzzy and Rough Sets, Inductive Logic Programming, Reinforcement Learning, Neural Networks and Support Vector Machines In addition, challenges in ML such as imbalanced data, perceptual computing, and pattern recognition of data which is episodic as well as temporal are also highlighted

Detection of Targets in Characteristic GPR Sensor Data Using Machine Learning Techniques

Abstract: Ground Penetrating Radar (GPR) is considered as one of the promising technologies to address the challenges of detecting buried threat objects, particularly in military applications. The aim of this research is to design a potential solution to the task of target detection and classification using GPR. This paper focuses on the first stage of this task which is target detection. Three approaches for automated target detection are presented; a probabilistic approach, an artificial neural network with direct data input, and an artificial neural network with frequency spaced features. These techniques are applied to a preliminary data set with promising results.

A multi-agent system architecture to classify colour images

Abstract: Colour image classification plays an important role in computer vision and pattern recognition. Traditional classification research mainly focuses on developing novel techniques that are efficient for image representation or classification. By processing considerable visual information, human can handle the complicated classification tasks quite effectively. Inspirited by the structure of the visual cortex, we propose a multi-agent colour image classification architecture (MACICA). Agents within a multi-agent system (MAS) architecture are programmed to deliver specific image classification capabilities. The MACICA provides an efficient classification output by sharing knowledge, communication and team work. The architecture is flexible and dynamic, while the platform has produced encouraging results, which are presented in the paper.

A Sequential Monte Carlo Approach for Brain Source Localization

Abstract: In this chapter we propose a solution to the Electroencephalography (EEG) inverse problem combining two techniques, which are the Sequential Monte Carlo (SMC) method for estimating the coordinates of the first two non-correlated dominative brain zones (represented by their respective current dipoles) and spatial filtering which is done by beamforming based on EEG data. Beamforming (BF) gives estimates of the respective source moments. In order to validate this novel approach for brain source localization, EEG data from dipoles with known locations and known moments are generated and artificially corrupted with noise. The noise represents the overall influence of other brain sources but they are not brain disturbances. When the power of the EEG signal due to the main brain sources is higher than the summed effect of all other secondary sources, the estimation of the localization of the leading sources is reliable and repetitive over a number of Monte Carlo runs.

Intelligent-Based Systems Engineering

Abstract: The International Council on Systems Engineering (INCOSE) defines Systems Engineering as an interdisciplinary approach and means to enable the realization of successful systems. Researchers are using intelligence-based techniques to support the practices of systems engineering in an innovative way. This research volume includes a selection of contributions by subject experts to design better systems.

Special issue: Computational intelligence models for image processing and information reasoning

Abstract: Computational Intelligence CI models comprise robust computing methodologies with a high level of machine learning quotient. CI models, in general, are useful for designing computerized intelligent systems/machines that possess useful characteristics mimicking human behaviors and capabilities in solving complex tasks, e.g., learning, adaptation, and evolution. Examples of some popular CI models include fuzzy systems, artificial neural networks, evolutionary algorithms, multi-agent systems, decision trees, rough set theory, knowledge-based systems, and hybrid of these models. This special issue highlights how different computational intelligence models, coupled with other complementary techniques, can be used to handle problems encountered in image processing and information reasoning.

Intelligent Interactive Multimedia Systems and Services: Proceedings of the 6th International Conference on Intelligent Interactive Multimedia Systems and Services (IIMSS2013)

Abstract: At a time when computers are more widespread than ever, intelligent interactive systems have become a necessity. The term multimedia systems refers to the coordinated storage, processing, transmission and retrieval of multiple forms of information, such as audio, image, video, animation, graphics and text. The growth of multimedia services has been exponential, as technological progress keeps up with the consumers need for content. The solution of 'one fits all' is no longer appropriate for the wide ranges of users with various backgrounds and needs, so one important goal of many intelligent interactive systems is dynamic personalization and adaptivity to users. This book presents 37 papers summarizing the work and new research results presented at the 6th International Conference on Intelligent Interactive Multimedia Systems and Services (KES-IIMSS2013), held in Sesimbra, Portugal, in June 2013. The conference series focuses on research in the fields of intelligent interactive multimedia systems and services and provides an internationally respected forum for scientific research in related technologies and applications.IOS Press is an international science, technical and medical publisher of high-quality books for academics, scientists, and professionals in all fields. Some of the areas we publish in: -Biomedicine -Oncology -Artificial intelligence -Databases and information systems -Maritime engineering -Nanotechnology -Geoengineering -All aspects of physics -E-governance -E-commerce -The knowledge economy -Urban studies -Arms control -Understanding and responding to terrorism -Medical informatics -Computer Sciences

The Systems Integration Technical Risk assessment fusing of Bayesian Belief Networks and Parametric Models

Abstract: This paper presents an approach for modelling Systems Integration Technical Risks SITR assessment using Bayesian Belief Networks BBN. SITR represent a significant part of project risks associated with a development of large software intensive systems. We propose conceptual modelling framework to address the problem of SITR assessment at early stages of a system life cycle. This framework includes a set of BBN models, representing the risk contributing factors, and complementing Parametric Models PM, used for providing input data to the BBN models. In particular we describe SITR identification approach explaining corresponding BBN models' topologies and relevant conceptual model framework. This framework includes a set of BBN models, representing the risk contributing factors, fused with complementary PMs providing input data to the BBN models. Heuristic approaches for easing Conditional Probabilities Tables CPT generation are described. We briefly discuss preliminary results of model testing. In conclusion we summarise benefits and constraints for SITR assessment based on BBN models, and provide suggestions for further research directions for model improvement.

Evaluating pilot situation awareness using multi-agent systems

Abstract: This paper presents the outcomes of collaborative research between three PhD students working in the Multi-Agent System MAS, Knowledge-Based System KBS, and aviation Situation Awareness SA domains. The aim of this research was to create a MAS that could be used to monitor pilot SA during flight. SA is a cognitive activity that is a critical function conducted by pilots to maintain knowledge of their environment during flight. Good SA ultimately enhances the safety of passengers by reducing the possibility of pilots contributing to a number of documented catastrophic incidents. A controlled experiment has been devised to enable these students to capture and analyse pilot behaviour in an attempt to passively monitor SA monitoring activities using a camera. The MAS consists of multiple agent capabilities that capture the pilots visual acuity and eye movements. This data is used to assess the perceived cognitive activity in real-time. All agents can communicate and share the knowledge captured in order to analyse the activity based on pattern-matching rules using an embedded KBS. The experiments confirmed it is possible to identify at least three specific behaviours. The agents were able to post-process the acquired data to distinguish significant differences between an expert pilot and trained volunteers.

Advances in Intelligent Signal Processing and Data Mining: Theory and Applications (Studies in Computational Intelligence

Abstract: Dealing with large amounts of data and reasoning in real time are some of the challenges that our every day life poses to us. The answer to these questions can be given by advanced methods in signal processing and data mining which is the scope of this book. The book presents theoretical and application achievements on some of the most efficient statistical and deterministic methods for information processing (filtering, clustering, decomposition, modelling) in order to extract targeted information and find hidden patterns. The techniques presented range from Bayesian approaches such as sequential Monte Carlo methods, Markov Chain Monte Carlo filters, Rao Blackwellization, to the biologically inspired paradigm of Neural Networks and decomposition techniques such as Empirical Mode Decomposition, Independent Component Analysis (ICA) and Singular Spectrum Analysis. Advances and new theoretical interpretations related with these techniques are detailed and illustrated on a variety of real life problems as multiple object tracking, group object tracking, localization in wireless sensor networks, brain source localization, behavior reasoning, classification, clustering, video sequence processing, and others.

Introduction to Intelligent Signal Processing and Data Mining

Abstract: Intelligent signal processing and data mining are the key components of present advances in many disciplines including science and engineering. This chapter presents a brief outline of contributions from some of the leading researchers in the field of intelligent signal processing and data mining.