Handbook of Research on
Computational Intelligencefor Engineering, Science,and Business
Siddhartha BhattacharyyaRCC Institute of Information Technology, India
Paramartha Dutta
Visva Bharati University, India
Volume I
Information Science
REFERENCE
Detailed Table of Contents
Preface xxvi
Acknowledgment xxx
Volume I
Section 1
Overview of Computational Intelligence
Chapter 1
Computational Intelligence Using Type-2 Fuzzy Logic Framework
A. Neogi, The University ofBurdwan, India
A. C. Mondal, The University ofBurdwan, India
S.K. Mandal, National Institute ofTechnical Teachers' Training & Research, India
In this chapter, the authors expand the notion oftype-2 fuzzy sets. An introduction to standard and interval
(type-2) fuzzy sets and systems is explained in the early part of the discussion. The chapter also covers
the ideas of hybrid type-2 fuzzy system. Next, the authors study the applicability of type-2 fuzzy logic(FL) system in student's performance in oral presentation as it is clearly new field ofresearch topic and
have an excellent opportunity to combine several fuzzy set method developed in the recent years. The
proposed application shows the linkage of type-2 fuzzy system with TOPSIS. The present chapter also
highlights the possible future directions for type-2 FL system research. By the end of the chapter, the
authors hope that even those with little previous experience of fuzzy logic should be enabled to applythese methods in their own application areas and/or begin research in this fascinating and exciting area.
Chapter 2
Soft Computing Based Statistical Time Series Analysis, Characterization of Chaos Theory, and
Theory ofFractals 30
Mofazzal H. Khondekar, Dr. B.C. Roy Engineering College, India
Dipendra N. Ghosh, Dr. B.C. Roy Engineering College, India
Koushik Ghosh, University Institute ofTechnology, University ofBurdwan, India
Anup Kumar Bhattacharya, National Institute ofTechnology, India
The present work is an attempt to analyze the various researches already carried out from the theoreti¬
cal perspective in the field of soft computing based time series analysis, characterization of chaos, and
theory of fractals. Emphasis has been given in the analysis on soft computing based study in prediction,
data compression, explanatory analysis, signal processing, filter design, tracing chaotic behaviour, and
estimation of fractal dimension of time series. The present work is a study as a whole revealing the ef¬
fectiveness as well as the shortcomings of the various techniques adapted in this regard.
Chapter 3
Machine Intelligence Using Hierarchical Memory Networks 62
A. P. James, IIITM-Kerala, India
This chapter presents the fundamentals ofa hardware based memory network that can perform complexcognitive tasks. The network is designed to provide space dimensionality reduction, which enables desired
functionality in a random environment. Complex network functionality is achieved by simple network
cells that minimize the needed chip area for hardware implementation. Functionality of this networkis demonstrated by automatic character recognition with various input deformations. In the character
recognition, the network is trained to recognize characters deformed by random noise, rotation, scal¬
ing, and shifting. This example demonstrates how cognitive functionality of a hardware network can be
achieved through an evolutionary process, as distinct from design based on mathematical formalism.
Section 2
Image Processing and Segmentation
Chapter 4
Image Analysis and Understanding Based on Information Theoretical Region MergingApproaches for Segmentation and Cooperative Fusion 75
Felipe Calderero, Universitat Pompeu Fabra (UPF), SpainFerran Marques, Technical University ofCatalonia (UPC), Spain
This chapter addresses the automatic creation of simplified versions of the image, known as imagesegmentation or partition, preserving the most semantically relevant information of the image at differ¬
ent levels of analysis. From a semantic and practical perspective, image segmentation is a first and key
step for image analysis and pattern recognition since region-based image representations provide a first
level of abstraction and a reduction of the number of primitives, leading to a more robust estimation of
parameters and descriptors. The proposed solution is based on an important class of hierarchical bottom-
up segmentation approaches, known as region merging techniques. These approaches naturally providea bottom-up hierarchy, more suitable when no a priori information about the image is available, and an
excellent compromise between efficiency of computation and representation. The chapter is organizedin two parts dealing with the following objectives: (i) provide an unsupervised solution to the segmen¬tation of generic images; (ii) design a generic and scalable scheme to automatically fuse hierarchical
segmentation results that increases the robustness and accuracy ofthe final solution.
Chapter 5
Multilevel Image Segmentation by a Multiobjective Genetic Algorithm Based OptiMUSlGActivation Function 122
Sourav De, The University ofBurdwan, India
Siddhartha Bhattacharyya, RCC Institute ofInformation Technology, India
Susanta Chakraborty, Bengal Engineering and Science University, India
The proposed chapter is intended to propose a self supervised image segmentation method by a multi-
objective genetic algorithm based optimized MUSIG (OptiMUSlG) activation function with a multi-
layer selforganizing neural network architecture to segment multilevel gray scale intensity images. The
multiobjective genetic algorithm based parallel version oftheOptiMUSIG (ParaOptiMUSIG) activation
function with a parallel selforganizing neural network architecture is also discussed to segment true color
images. These methods are quite efficient enough to overcome the drawbacks of the single objectivebased OptiMUSIG and ParaOptiMUSIG activation functions to segment gray scale and true color images,respectively. The proposed multiobjective genetic algorithm based optimization methods are applied on
three standard objective functions to measure the quality ofthe segmented images. These functions form
the multiple objective criteria ofthe multiobjective genetic algorithm based image segmentation method.
Chapter 6
A Novel Fuzzy Rule Guided Intelligent Technique for Gray Image Extraction and Segmentation ..163
Koushik Mondal, IITIndore, India
Image segmentation and subsequent extraction from a noise-affected background, has all along remaineda challenging task in the field of image processing. There are various methods reported in the literature
to this effect. These methods include various Artificial Neural Network (ANN) models (primarily super¬
vised in nature), Genetic Algorithm (GA) based techniques, intensity histogram based methods, etcetera.
Providing an extraction solution working in unsupervised mode happens to be even more interesting a
problem. Fuzzy systems concern fundamental methodology to represent and process uncertainty and
imprecision in the linguistic information. The fuzzy systems that use fuzzy rules to represent the domain
knowledge of the problem are known as Fuzzy Rule Base Systems (FRBS). Literature suggests that
effort in this respect appears to be quite rudimentary. This chapter proposes a fuzzy rule guided novel
technique that is functional devoid ofany external intervention during execution. Experimental results
suggest that this approach is an efficient one in comparison to different other techniques extensivelyaddressed in literature. In order to justify the supremacy of performance of our proposed technique in
respect of its competitors, the author takes recourse to effective metrices like Mean Squared Error (MSE).Mean Absolute Error (MAE), and Peak Signal to Noise Ratio (PSNR).
Chapter 7
Graph Based Segmentation of Digital Images 182
B.K. Tripathy, VIT University, India
P. V.S.S.R. Chandra Mouli, VIT University, India
Image Segmentation is the process of dividing an image into semantically relevant regions. The problemis still an active area due to wide applications in object detection and recognition, image retrieval, im¬
age classification, et cetera. The problem is challenging due to its subjective nature. Many researchers
addressed this problem by exploring graph theoretic principles. The key idea is the transformation of
segmentation problem into graph partitioning problem by representing the image as a graph. The aimof this chapter is to study various graph based segmentation algorithms.
Chapter 8
Development of a Stop-Line Violation Detection System for Indian Vehicles 200
Satadal Saha, MCKV Institute ofEngineering, India
Subhadip Basu, Jadavpur University, India
Mita Nasipuri, Jadavpur University, India
In the present work, the authors designed and developed a complete system for generating the list ofall
violating vehicles that has violated the stop-line at a road crossing automatically from video snapshotsof road-side surveillance cameras using background subtraction technique. It then localizes the license
plates of the vehicles by analyzing the vertical edge map of the images, segments the license plate
characters using connected component labeling algorithm, and recognizes the characters using back
propagation neural network. Considering round-the-clock operations in a real-life test environment, the
developed system could successfully track 92% images of vehicles with violations on the stop-line in a
red traffic signal. The performance of the system is evaluated with a dataset of 4717 images collected
from 13 different camera views in 4 different environmental conditions. The authors have achieved
around 92% plate localization accuracy over different views and weather conditions. The average platelevel recognition accuracy of 92.75% and character level recognition accuracy of 98.76% are achieved
over the localized vehicle images.
Chapter 9
A Comparative Study of Unsupervised Video Shot Boundary Detection Techniques UsingProbabilistic Fuzzy Entropy Measures 228
Biswanath Chakraborty, RCC Institute ofInformation Technology, India
Siddhartha Bhattacharyya. RCC Institute ofInformation Technology, India
Susanta Chakraborty, Bengal Engineering and Science University, India
The performance of video shot boundary detection technique in unsupervised video sequence can be
improved by the use of different probabilistic fuzzy entropies. In this chapter, the authors present a new
technique for identifying as to whether there are any appreciable changes from one video context to
another in the available sequence of image frames extracted from a mixture of a numbers of video files.
They then compared their technique with an existing technique and found improved performance ofthevideo shot boundary detection techniques using probabilistic fuzzy entropies.
Chapter 10
A Hierarchical Multilevel Image Thresholding Method Based on the Maximum Fuzzy Entropy
Principle 241
Pearl P. Guan, City University ofHong Kong, Hong Kong
Hong Yan. City University of Hong Kong, Hong Kong & University ofSydney, Australia
Image thresholding and edge detection are crucial in image processing and understanding. In this chapter,the authors propose a hierarchical multilevel image thresholding method for edge information extraction
based on the maximum fuzzy entropy principle. In orderto realize multilevel thresholding, a tree structure
is used to express the histogram of an image. In each level of the tree structure, the image is segmentedby three-level thresholding based on the maximum fuzzy entropy principle. In theory, the histogram
hierarchy can be combined arbitrarily with multilevel thresholding. The proposed method is proven by
experimentation to retain more edge information than existing methods employing several grayscaleimages. Furthermore, the authors extend the multilevel thresholding algorithm for color images in the
application of content-based image retrieval, combining with edge direction histograms. Compared to
using the original images, experimental results show that the thresholding images outperform in achiev¬
ing higher average precision and recall.
Chapter 11
Adaptive Median Filtering Based on Unsupervised Classification of Pixels 273
./. K. Mandal, University ofKalyani. India
Somnath Mukhopadhyay, Aryabhatta Institute ofEngineering & Management Durgapur, India
This chapter deals with a novel approach which aims at detection and filtering of impulses in digital
images through unsupervised classification of pixels. This approach coagulates directional weightedmedian filtering with unsupervised pixel classification based adaptive window selection toward detec-
tion and filtering of impulses in digital images. K-means based clustering algorithm has been utilized
to detect the noisy pixels based adaptive window selection to restore the impulses. Adaptive median
filtering approach has been proposed to obtain best possible restoration results. Results demonstratingthe effectiveness ofthe proposed technique are provided for numeric intensity values described in terms
offeature vectors. Various benchmark digital images are used to show the restoration results in terms of
PSNR (dB) and visual effects which conform better restoration of images through proposed technique.
Section 3
Database Oriented Techniques
Chapter 12
Data Clustering Algorithms Using Rough Sets 297
B.K. Tripathy, VIT University, India
Adhir Ghosh, VIT University, India
Developing Data Clustering algorithms have been pursued by researchers since the introduction of
k-means algorithm (Macqueen 1967; Lloyd 1982). These algorithms were subsequently modified to
handle categorical data. In order to handle the situations where objects can have memberships in multiple
clusters, fuzzy clustering and rough clustering methods were introduced (Lingras et al 2003, 2004a).
There are many extensions of these initial algorithms (Lingras et al 2004b; Lingras 2007; Mitra 2004;
Peters 2006,2007). The MMRalgorithm (Parmar et al 2007), its extensions (Tripathy et al 2009,2011 a,
201 lb) and the MADE algorithm (Herawan et al 2010) use rough set techniques for clustering. In this
chapter, the authors focus on rough set based clustering algorithms and provide a comparative study of
all the fuzzy set based and rough set based clustering algorithms in terms of their efficiency. They also
present problems for future studies in the direction of the topics covered.
Chapter 13
Evolution of Genetic Algorithms in Classification Rule Mining 328
Dipankar Dutta, University Institute of Technology, The University ofBurdwan, India
Jaya Sil, Bengal Engineering and Science University, India
Classification is one of the most studied areas of data mining, which gives classification rules duringtraining or learning. Classification rule mining, an important data-mining task, extracts significant rules
for classification of objects. In this chapter class specific rules are represented in IF <Antecedent>
THEN <Consequent> form. With the popularity of soft computing methods, researchers explore dif¬
ferent soft computing tools for rule discovery. Genetic algorithm (GA) is one of such tools. Over time,new techniques of GA for forming classification rules are invented. In this chapter, the authors focus
on an understanding of the evolution of GA in classification rule mining to get an optimal rule set that
builds an efficient classifier.
Chapter 14
Database Anonymization Techniques with Focus on Uncertainty and Multi-Sensitive Attributes...
364
B. K. Tripathy, VIT University, India
Publication of Data owned by various organizations for scientific research has the danger of sensitive
information of respondents being disclosed. The policy of removal or encryption of identifiers cannot
avoid the leakage of information through quasi-identifiers. So, several anonymization techniques like
k-anonymity, 1-diversity, and t-closeness have been proposed. However, uncertainty in data cannot be
handled by these algorithms. One solution to this is to develop anonymization algorithms by using roughset based clustering algorithms like MMR, MMeR, SDR, SSDR, and MADE at the clustering stage of
existing algorithms. Some ofthese algorithms handle both numerical and categorical data. In this chap¬ter, the author addresses the database anonymization problem and briefly discusses k-anonymizationmethods. The primary focus is on the algorithms dealing with 1-diversity of databases having single or
multi-sensitive attributes. The author also proposes certain algorithms to deal with anonymization of
databases with involved uncertainty. Also, the aim is to draw attention of researchers towards the vari¬
ous open problems in this direction.
Chapter 15
Using Data Masking for Balancing Security and Performance in Data Warehousing 384
Ricardo Jorge Santos, CISUC - FCTUC - University ofCoimbra, Portugal
Jorge Bernardino, CISUC - ISEC - Polytechnic Institute ofCoimbra, PortugalMarco Vieira, CISUC -FCTUC - University ofCoimbra, Portugal
Data Warehouses (DWs) are the core ofsensitive business information, which makes them an appealingtarget. Encryption solutions are accepted as the best way to ensure strong security in data confidentialitywhile keeping high database performance. However, this work shows that they introduce massive stor¬
age space and performance overheads to a magnitude that makes them unfeasible for DWs. This work
proposes a data masking technique for protecting sensitive business data in DWs which balances securitystrength with database performance, using a formula based on the mathematical modular operator and
simple arithmetic operations. The proposed solution provides apparent randomness in the generationand distribution of the masked values, while introducing small storage space and query execution time
overheads. It also enables a false data injection method for misleading attackers and increasing the
overall security strength. It can be easily implemented in any DataBase Management System (DBMS)and transparently used, without changes to application source code. Experimental evaluations using a
real-world DWand TPC-H decision support benchmark implemented in leading DBMS Oracle 11 g and
Microsoft SQL Server 2008 demonstrate its overall effectiveness. Results show the substantial savingsof its implementation costs when compared with state of the art encryption solutions provided by those
DBMS and that it outperforms those solutions in both data querying and insertion of new data.
Volume II
Chapter 16
Schedulers Based on Ant Colony Optimization for Parameter Sweep Experiments in
Distributed Environments 410
Elina Pacini, Institutefor Information and Communication Technologies, Universidad
Nacional de Cuyo, ArgentinaCristian Mateos, Instituto Superior de Ingenieria de Software, Consejo Nacional de
Investigaciones Cientificas y Tecnicas, ArgentinaCarlos Garcia Garino, Institute for Information and Communication Technologies,
UniversidadNacional de Cuyo, Argentina
Scientists and engineers are more and more faced to the need ofcomputational power to satisfy the ever-
increasing resource intensive nature of their experiments. An example ofthese experiments is Parameter
Sweep Experiments (PSE). PSEs involve many independentjobs, since the experiments areexecuted under
multiple initial configurations (input parameter values) several times. In recent years, technologies such
as Grid Computing and Cloud Computing have been used for running such experiments. However, for
PSEs to be executed efficiently, it is necessary to develop effective scheduling strategies to allocatejobs
to machines and reduce the associated processing times. Broadly, the job scheduling problem is known
to be NP-complete, and thus many variants based on approximation techniques have been developed. In
this work, the authors conducted a survey ofdifferent scheduling algorithms based on Swarm Intelligence
(SI), and more precisely Ant Colony Optimization (ACO), which is the most popular SI technique, to
solve the problem ofjob scheduling with PSEs on different distributed computing environments.
Section 4
Classification, Design, and Modeling
Chapter 17
Particle Swarm Optimization Algorithm and its Hybrid Variants for Feature Subset Selection 449
Basabi Chakraborty, Iwate Prefectural University, Japan
Selecting an optimum subset of features from a large set of features is an important pre- processing step
forpattern classification, data mining, or machine learning applications. Feature subset selection basically
comprises of defining a criterion function for evaluation of the feature subset and developing a search
strategy to find the best feature subset from a large number of feature subsets. Lots of mathematical and
statistical techniques have been proposed so far. Recently biologically inspired computing is gaining
popularity for solving real world problems for their more flexibility compared to traditional statistical
or mathematical techniques. In this chapter, the role of Particle Swarm Optimization (PSO), one of the
recently developed bio-inspired evolutionary computational (EC) approaches in designing algorithmsfor producing optimal feature subset from a large feature set, is examined. A state of the art review
on Particle Swarm Optimization algorithms and its hybrids with other soft computing techniques for
feature subset selection are presented followed by author's proposals of PSO based algorithms. Simple
simulation experiments with benchmark data sets and their results are shown to evaluate their respec¬
tive effectiveness and comparative performance in selecting best feature subset from a set of features.
Chapter 18
An Evolving System in the Text Classification Problem 467
Elias Oliveira, Universidade Federal do Espirito Santo, Brazil
Patrick Marques CiareIJi, Universidade Federal do Espirito Santo, Brazil
Evandro Ottoni Teatini Salles, Universidade Federal do Espirito Santo. Brazil
Traditional machine learning techniques have been successful in yielding good results when the data
are stable along the time horizon. However, in many cases, these techniques may be inefficient for data
that are constantly expanding and changing over time. To address this problem, new learning techniques
have been proposed in the literature. In this chapter, the authors discuss some improvements on their
technique, called Evolving Probabilistic Neural Network (ePNN), and present the aspects of this recent
learning paradigm. This technique is based on the Probabilistic Neural Networks. In this chapter the au¬
thors compare their technique against two other competitive techniques that can be found in the literature:
Incremental Probabilistic Neural Network (IPNN) and Evolving Fuzzy Neural Network (EFuNN). To
show the better performance of their technique, the authors present and discuss a series ofexperimentsthat demonstrate the efficiency of ePNN over both the IPNN and EFuNN approaches.
Chapter 19
Fuzzy Based Modeling, Control, and Fault Diagnosis of Permanent Magnet
Synchronous Generator 487
N. Selvaganesan, Indian Institute ofSpace Science and Technology (IIST), Govt, ofIndia. India
This chapter presents the design methodology of fuzzy based modeling, control, and fault diagnosisof Permanent Magnet Synchronous Generator (PMSG) system. The fuzzy based modeling scheme for
PMSG is developed using the general Takagi-Sugeno fuzzy model. Subsequently, fuzzy controller is
designed and simulated to maintain three phase output voltage as constant by controlling the speed of
generator. The feasibility ofthe fuzzy model and control scheme is demonstrated using various operatingconditions by MATLAB simulation. Also, fuzzy based fault detection scheme for PMSG is developedand presented. The positive and negative sequence currents are used as fault indicators and given as
inputs to fuzzy fault detector. The fuzzy inference system is created, and rule base is evaluated, relatingthe sequence current component to the type of faults. The feasibility of this scheme is demonstrated for
different types of fault under various operating conditions using MATLAB/Simulink.
Chapter 20
Algorithms and Principles for Intelligent Design of Flapping Wing Micro Aerial Vehicles 521
Ajay Bangalore Harish. Indian Institute ofScience, India
Dineshkumar Harursampath, Indian Institute ofScience. India
Almost all Micro Aerial Vehicles (MAVs) designed so far facilitate the flapping motion oftheir wings bymeans of a mounted actuating mechanism, driven, for example, by a piezoelectric crystal. The develop¬ments over the past decade or so in smart material technologies like the invention of Piezoelectric Fiber
Reinforced Composite (PFRC) materials and innovative manufacturing techniques to reduce cost have
resulted in favorable materials for dynamic actuating applications. Thus, the concept ofactively deform-
able wings to produce combined flapping and feathering actions is evolving as an attractive enabler for
design of future MAVs. A smart material like PFRC can both sense and actuate in a collocated fashion,thus building an additional level of computational intelligence into the MAV itself. Such a promisingopportunity indicates an urgent need for reliable design tools to accelerate development of MAVs. In
this work, the authors propose a modular design tool specifically for design of self-actuating flappingwing MAVs.
Chapter 21
Fuzzy-Controlled Energy-Conservation Technique (FET) for Mobile ad hoc Networks 556
Anuradha Banerjee. Kalyani Government Engineering College. India
Nodes in ad hoc networks have limited battery power; hence, they require energy efficient techniquesto improve average node lifetime and network performance. Maintaining energy efficiency in network
communication is really challenging because highest energy efficiency is achieved if all the nodes are
switched off and maximum network throughput is obtained if all the nodes are fully operational, i.e.
always turned on. A promising energy conservation technique for the ad hoc networks must maintain
effective packet forwarding capacity while turning off the network interface of very busy nodes for
some time and redirecting the traffic through some comparatively idle nodes roaming around them. This
also helps in fair load distribution in the network and maintenance of network connectivity by reducingthe death rate (complete exhaustion of nodes). The present chapter proposes a fuzzy-controlled energyconservation technique (FET) that identifies the busy and idle nodes to canalize some traffic of busynodes through the idle ones. In simulation section, the FET embedded versions of several state-of-the-
art routing protocols in ad hoc networks are compared with their ordinary versions and the results quite
emphatically establish the superiority of FET-embedded versions in terms of packet delivery ratio, mes¬
sage cost, and network energy consumption. End-to-end delay also reduces significantly.
Chapter 22
Decision Fusion of Multisensor Images for Human Face Identification in Information Security 571
Mrinal Kanti Bhowmik, Tripura University, India
Priya Saha, Tripura University, India
Goutam Majumder, Tripura University, India
Debotosh Bhattacharjee, Jadavpur University, India
The chapter is mainly focused on theoretical discussions and experimental observations on decision
fusion along with feature level multisensor fusion technique for human face identification especiallyuseful in information security system in order to obtain better recognition rate. Feature level multisensor
fusion ofoptical and infrared images is performed to resolve the difficulties of individual interpretationof visual and infrared images as a first step of the face identification system. ROC curve analysis is also
reported to verify the recognition accuracy after classification offused images using Support Vector Ma¬
chine (SVM). The authors have performed experiments on IRIS face database in three different groups:full dataset, and two subsets with variation in expression and illumination. Classification accuracy was
obtained in two subsets and full dataset as 95%, 94.12%, and 99.08%, respectively after decision fusion.
Chapter 23
Modernization of Healthcare and Medical Diagnosis System Using Multi Agent System (MAS):A Comparative Study 592
Shibakali Gupta, University Institute ofTechnology, Burdwan University, India
Sripati Mukherjee, Burdwan University, India
Sesa Singha Roy, Tata Consultancy Service, India
The healthcare system that prevailed some years ago was a mere pen and paper based system. A number
ofworkers, staff, and written records were the main components ofthe prevailing system of healthcare.
This had a number of drawbacks, and a number of mishaps occurred due to mismanagement ofdata andinformation. There was a need for development. Then, the concept oftelemedicine came, which revolu¬
tionized the healthcare paradigm to a great extent. With the advancement of telemedicine, many majorproblems of the prevailing system were removed. But, still there were many other aspects which could
be further improved to make healthcare facilities more enhanced. Keeping this in mind, the concept of
Multi Agent System (MAS) was introduced in the healthcare system later. MASes are considered as the
best and most appropriate technology that can be used in the development of applications in healthcare
paradigm where the presence of multiple agents, heterogeneous and loosely coupled components, the
data management in a dynamic and distributed environment, and multi-user collaborations are consid¬
ered the most pertinent requirements for healthcare system. This chapter focuses mainly about MAS,its applications, and some systems that were developed by the authors.
Section 5
Applications
Chapter 24
Watermarking of Data Using Biometrics 623
Swanirbhar Majumder, Deemed University, India
Tirtha Sankar Das, RCC Institute ofInformation Technology, India
These days, for the copyright protection and security of multimedia data in this age of the tech-savvyworld, watermarking is a very important technique. Moreover, with the inclusion of biometrics for the
watermarking schemes, the concept of "something you are" is included in the watermark and/or cover
image. This thereby increases the security intensity in the multimedia data. And to give a glimpse ofthe
technique the concepts of Watermarking, biometric and watermarking using biometrics is discussed.
Finally, a particular case of real time watermarking of data using biometric is discussed by specifyinga practical example.
Chapter 25
Quantum Backpropagation Neural Network Approach for Modeling of Phenol Adsorptionfrom Aqueous Solution by Orange Peel Ash 649
Siddhartha Bhattacharjee, Tata Consultancy Services, India
Siddhartha Bhattacharyya, RCC Institute ofInformation Technology, India
Naba Kumar Mondal, The University ofBurdwan, India
The chapter describes a multilayer quantum backpropagation neural network (QBPNN) architecture
to predict the removal of phenol from aqueous solution by orange peel ash, guided by the applicationof three types of activation functions and characterized by backpropagation of errors. These activation
functions are Sigmoid function, tanh function and tan 1.5h function. First by a classical multilayer neural
network architecture with three types ofactivation functions is discussed in this chapter. It takes 6000000
iterations to train the network with a learning rate of 0.01. Among these three types of activation func¬
tions tan 1.5 function shows the best prediction result. Next, QBPNN is discussed in this chapter. It takes
22000 iterations to train the network with the same learning rate. Here also tanl.Sh function shows the
best result in prediction of removal of phenol. Thus QBPNN is much faster than the classical multilayerneural network architecture. Different graphs are also given for comparison between the experimentaloutput and network output using different activation functions. This particular chapter basically deals witha model application by which experimental results can be comparing with the model output. Because of
their reliable, robust, and salient characteristics in capturing the non-linear relationships existing between
variables (multi-input/output) in complex systems, it has become apparent that numerous applications of
ANNs/QBNN have been successfully conducted in various parts of environmental engineering. Fuzzy
Logic is also used as alternate method to predict the removal of phenol from aqueous solution by orange
peel ash, but QBPNN shows the best result.
Chapter 26
Computational Intelligence for Pathological Issues in Precision Agriculture 672
Sanjeev S. Sannakki, Gogte Institute ofTechnology, India
Vijay S. Rajpurohit, Gogte Institute of Technology, India
V. B. Nargund, University ofAgricultural Sciences, India
Amn R. Kumar, Ashokrao Mane Group ofInstitutions, India
Prema S. Yallur, Ashokrao Mane Group ofInstitutions, India
Plant Pathology is the scientific study of plant diseases, caused by pathogens and environmental condi¬
tions (physiological factors). Detection and grading of plant diseases by machine vision is an essential
research topic as it may prove useful in monitoring large fields of crops. This can be of great benefit to
those users, who have little or no information about the crop they are growing. Also, in some develop¬
ing countries, farmers may have to go long distances to contact experts to dig up information which is
expensive and time consuming. Therefore, looking for a fast, automatic, less expensive, and accurate
method to detect plant diseases is of great realistic significance. Such an efficient system can be mod¬
eled by integrating the various tools/techniques of information and communication technology (ICT)in agriculture. The objective of the present chapter is to model an intelligent decision support systemfor detection and grading of plant diseases which encompasses image processing techniques and soft
computing/machine learning techniques.
Chapter 27
Cancer Gene Expression Data Analysis Using Rough Based Symmetrical Clustering 699
Anasua Sarkar, Government College ofEngineering and Leather Technology, India
Ujjwal Maulik, Jadavpur University, India
Identification of cancer subtypes is the central goal in the cancer gene expression data analysis. Modi¬
fied symmetry-based clustering is an unsupervised learning technique for detecting symmetrical convex
or non-convex shaped clusters. To enable fast automatic clustering of cancer tissues (samples), in this
chapter, the authors propose a rough set based hybrid approach for modified symmetry-based clusteringalgorithm. A natural basis for analyzing gene expression data using the symmetry-based algorithm is
to group together genes with similar symmetrical patterns of microarray expressions. Rough-set theory
helps in faster convergence and initial automatic optimal classification, thereby solving the problemof unknown knowledge of number of clusters in gene expression measurement data. For rough-set-theoretic decision rule generation, each cluster is classified using heuristically searched optimal reducts
to overcome overlapping cluster problem. The rough modified symmetry-based clustering algorithm is
compared with another newly implemented rough-improved symmetry-based clustering algorithm and
existing K-Means algorithm over five benchmark cancer gene expression data sets, to demonstrate its
superiority in terms ofvalidity. The statistical analyses are also performed to establish the signifizzvcanceof this rough modified symmetry-based clustering approach.
Chapter 28
Computer Intelligence in Healthcare 716
Pramit Ghosh, RCC Institute ofInformation Technology, India
Debotosh Bhattacharjee, Jadavpur University, India
Mita Nasipuri, Jadavpur University, India
Dipak Kumar Basu, Jadavpur University, India
Low cost solutions for the development of intelligent bio-medical devices that not only assist peopleto live in a better way but also assist physicians for better diagnosis are presented in this chapter. Twosuch devices are discussed here, which are helpful for prevention and diagnosis of diseases. Statistical
analysis reveals that cold and fever are the main culprits for the loss ofman-hours throughout the world,and early pathological investigation can reduce the vulnerability ofdisease and the sick period. To reduce
this cold and fever problem a household cooling system controller, which is adaptive and intelligent in
nature, is designed. It is able to control the speed of a household cooling fan or an air conditioner based
on the real time data, namely room temperature, humidity, and time for which system is active, which are
collected from environment. To control the speed in an adaptive and intelligent manner, an associative
memory neural network (Kramer) has been used. This embedded system is able to learn from trainingset; i.e., the user can teach the system about his/her feelings through training data sets. When the system
starts up, it allows the fan to run freely at full speed, and after certain interval, it takes the environmental
parameters like room temperature, humidity, and time as inputs. After that, the system takes the decision
and controls the speed of the fan.
Chapter 29
Intelligence in Web Technology 739
Sourav Mailra, Burdwan University, India
A. C. Mondal, Burdwan University, India
End users also start days with Internet. This has become the scenario. One of the most burgeoning needs
of computer science research is research on web technologies and intelligence, as that has become
one of the most emerging nowadays. A big area of other research areas like e-marketing, e-leaming,e-governance, searching technologies, et cetera will be highly benefited if intelligence can be added to
the Web. The objective of this chapter is to create a clear understanding of Web technology research and
highlight the ways to implement Semantic Web. The chapter also discusses the tools and technologiesthat can be applied to develop Semantic Web. This new research area needs enough care as sometimes
data are open. Thus, software engineering issues are also a focus.
Chapter 30
Performance Comparison of Different Intelligent Techniques Applied on Detecting Proportionof Different Component in Manhole Gas Mixture 758
Varun Kumar Ojha, Visva Bharati University, India
Paramartha Dutta, Visva Bharati University, India
This chapter deals with the comparison of performances ofdifferent intelligent techniques for detectingproportion ofdifferent component gases present in manhole gas mixture. Toxic gases found in manhole
gas mixture are Hydrogen Sulfide (H2S). Ammonia (NH3). Methane (CH4), Carbon Dioxide (C02),Nitrogen Oxide (NOx), Carbon Monoxide (CO), et cetera. Detection of these toxic gases is essential
since these gases influence human health even due to very short exposure. This study is centered on
design issues of an intelligent sensory system for detecting proportion of different components in man¬
hole gas mixture and comparison of different intelligent techniques applied for this. The investigationencompasses linear regression based statistical technique, backpropagation algorithm, neuro genetictechniques (using genetic algorithm to train neural network), and neuro swarm techniques (using particleswarm optimization algorithm to train neural network).
Compilation of References xxxi
About the Contributors ciii
Index cxvii