*Call for Chapters: Intelligent Multidimensional Data Clustering and
Analysis*
Editors
Dr. Siddhartha Bhattacharyya, RCC Institute of Information Technology, India
Dr. Sourav De, University Institute of Technology, India
Dr. Indrajit Pan, RCC Institute of Information Technology, India
Prof. (Dr.) Paramartha Dutta, Visva-Bharati University, India
Call for Chapters
Please click
http://www.igi-global.com/publish/call-for-papers/call-details/1849 to
submit a chapter proposal.
Proposals Submission Deadline: November 30, 2015
Full Chapters Due: January 30, 2016
Introduction
Commonly used as a preliminary data mining practice, data preprocessing
transforms the data into a format that will be more easily and effectively
processed for the purpose of the users. There are a number of data
preprocessing techniques: data cleaning, data integration, data
transformation and data reduction. The need to cluster large quantities of
multi-dimensional data is widely recognized. Cluster analysis is used to
identify homogeneous and well-separated groups of objects in databases. It
plays an important role in many fields of business and science.
Existing clustering algorithms can be broadly classified into four types:
partitioning, hierarchical, grid-based, and density-based algorithms.
Partitioning algorithms start with an initial partition and then use an
iterative control strategy to optimize the quality of the clustering
results by moving objects from one group to another.
Hierarchical algorithms create a hierarchical decomposition of the given
data set of data objects. Grid-based algorithms quantize the space into a
finite number of grids and perform all operations on this quantized space.
Density-based approaches are designed to discover clusters of arbitrary
shapes. These approaches hold that, for each point within a cluster, the
neighborhood of a given radius must exceed a defined threshold. Each of the
existing clustering algorithms has both advantages and disadvantages. The
most common problem is rapid degeneration of performance with increasing
dimensions, particularly with approaches originally designed for
low-dimensional data. To solve the high-dimensional clustering problem,
dimension reduction methods have been proposed which assume that clusters
are located in a low-dimensional subspace. However, this assumption does
not hold for many real-world data sets. The difficulty of high-dimensional
clustering is primarily due to the following characteristics of
high-dimensional data:
High-dimensional data often contain a large amount of noise (outliers). The
existence of noise results in clusters which are not well-separated and
degrades the effectiveness of the clustering algorithms. Clusters in
high-dimensional spaces are commonly of various densities. Grid-based or
density-based algorithms therefore have difficulty choosing a proper cell
size or neighborhood radius which can find all clusters. Clusters in
high-dimensional spaces rarely have welldefined shapes, and some algorithms
assume clusters of certain shapes. The effectiveness of grid-based
approaches suffer when data points are clustered around a vertex of the
grid and are separated in different cells.
To sum up, the classical techniques lack in one way or other as regards to
faithful analysis and clustering of multidimensional data owing to inherent
uncertainties in assumptions and heuristic choices. It is in this scenario
that the soft computing paradigm can be effectively used to arrive at
effective and productive throughputs.
Objective
To bring a broad spectrum of multidimensional data clustering and data
analysis applications under the purview of hybrid intelligence so that it
is able to trigger further inspiration among various research communities
to contribute in their respective fields of applications thereby orienting
these application fields towards intelligence. Once the purpose, as stated
above, is achieved a larger number of research communities may be brought
under one umbrella to ventilate their ideas in a more structured manner. In
that case, the present endeavor may be seen as the beginning of such an
effort in bringing various research applications close to one another. By
academically coming closer to one another, research communities working in
diversified application areas involving multidimensional data viz. true
color images, videos, big data, would be more encouraged to form groups
among themselves paving way for interdisciplinary research. Speaking from
the scholastic view, this is a formidable achievement in which the present
endeavor may be thought of as the maiden facilitator. It may however be
noted that there are good amounts of contributions of the application of
hybrid soft computing in various fields. However, any such previous effort
has remained application specific i. e. aimed at identifying a specific
application domain where the ingredients of hybrid soft computing have been
applied quite effectively. But, to the best of our knowledge, efforts to
bring in multiple domains of multidimensional data within one framework are
not very frequent. In that sense, this appears to be the first such effort
to accommodate cross platform applications of hybrid soft computing.
Moreover, efforts of hybridization are very meager in the literature. Once
successful, this will become an encouragement towards further research of
interdisciplinary nature by providing scope to various research communities
to come together through such an effort.
Target Audience
The proposed book would come to the benefits of several categories of
students and researchers. At the students level, this book can serve as a
treatise/reference book for the special papers at the masters level aimed
at inspiring possibly future researchers. Newly inducted PhD aspirants
would also find the contents of this book useful as far as their compulsory
courseworks are concerned. At the researchers' level, those interested in
interdisciplinary research would also be benefited from the book. After
all, the enriched interdisciplinary contents of the book would always be a
subject of interest to the faculties, existing research communities and new
research aspirants from diverse disciplines of the concerned departments of
premier institutes across the globe. This is expected to bring different
research backgrounds (due to its cross platform characteristics) close to
one another to form effective research groups all over the world. Above
all, availability of the book should be ensured to as much universities and
research institutes as possible through whatever graceful means it may be.
Recommended Topics
PART - I: Theoretical Foundation of Hybrid Intelligence:
Computational Intelligence: foundations and principles; neural networks;
fuzzy systems; near set; soft set; evolutionary computation; rough sets;
swarm intelligence
Hybridization of intelligent techniques
PART - II: Hybrid Soft Computing Paradigm:
Algorithmic, experimental, prototyping and implementation
Neuro-Fuzzy, Neuro-genetic, Fuzz-genetic, Neuro-fuzz-genetic
architectures etc.
Rough-fuzzy, Rough-neuro, Rough-neuro-fuzz architectures and the like
Quantum inspired hybrid soft computing architectures
PART - III: Introduction to Multidimensional Data
Types of Data Sets, Images, Videos, Big Data, Homogeneous and
Heterogeneous data
Characteristics of Data Sets
Common Errors in Data Sets, Missing Values
Outliers
Data Correlation
Standard Data Sets
PART - IV: Data Preprocessing
Expert Knowledge
Outlier Removal
Noise Removal
Handling of missing values
Distance Metrics
Data Normalization
PART - V: Characterizing the uncertainty
Types of uncertainty, methods for handling the uncertainty
Bootstrap confidence intervals of the phenomenon
Permutation tests
Randomization
PART VI: Data Handling Methods
Dimensionality Reduction
Singular Valued Decomposition
Noise Removal
Principal Component Analysis
Independent Component Analysis
Feature Selection
Segmentation
PART - VI: Data Clustering
K-means clustering
k-nearest neighbor clustering
fc-means clustering
Hierarchical clustering
Support Vector Machines
Decision Trees
Automatic Data Clustering Algorithms
PART - VII: Data Analysis
Association Rules
Multilevel Image Segmentation
Video Segmentation
Rough-Fuzzy Data Analysis
PART - VIII: Data Mining
Models - Supervised, Unsupervised
Supervised - Classification, Regression
Unsupervised - Clustering, Latent variable models
Cross validation
Submission Procedure
Researchers and practitioners are invited to submit on or before November
30, 2015, a chapter proposal of 1,000 to 2,000 words clearly explaining the
mission and concerns of his or her proposed chapter. Full chapters are
expected to be submitted by January 30, 2016. All submitted chapters will
be reviewed on a double-blind review basis. Contributors may also be
requested to serve as reviewers for this project.
Publisher
This book is scheduled to be published by IGI Global (formerly Idea Group
Inc.), publisher of the "Information Science Reference" (formerly Idea
Group Reference), "Medical Information Science Reference," "Business
Science Reference," and "Engineering Science Reference" imprints. For
additional information regarding the publisher, please visit
www.igi-global.com. This publication is anticipated to be released in 2017.
Important Dates
November 30, 2015: Proposal Submission Deadline
January 30, 2016: Full Chapter Submission
February 28, 2016: Revised Chapter Submission
Inquiries
Dr. Siddhartha Bhattacharyya
Department of Information Technology
RCC INSTITUTE OF INFORMATION TECHNOLOGY
CANAL SOUTH ROAD, BELIAGHATA, KOLKATA – 700 015, INDIA
M: +919830354195
E-mail: [log in to unmask]
