The problem of classification is perhaps one of the most widely studied in the data mining and machine learning communities. This problem has been studied by researchers from several disciplines
over several decades. Applications of classification include a wide variety of problem domains such
as text, multimedia, social networks, and biological data. Furthermore, the problem may be encountered in a number of different scenarios such as streaming or uncertain data. Classification is a
rather diverse topic, and the underlying algorithms depend greatly on the data domain and problem
scenario.
Therefore, this book will focus on three primary aspects of data classification. The first set of
chapters will focus on the core methods for data classification. These include methods such as probabilistic classification, decision trees, rule-based methods, instance-based techniques, SVM methods, and neural networks. The second set of chapters will focus on different problem domains and
scenarios such as multimedia data, text data, time-series data, network data, data streams, and uncertain data. The third set of chapters will focus on different variations of the classification problem
such as ensemble methods, visual methods, transfer learning, semi-supervised methods, and active
learning. These are advanced methods, which can be used to enhance the quality of the underlying
classification results.
The classification problem has been addressed by a number of different communities such as
pattern recognition, databases, data mining, and machine learning. In some cases, the work by the
different communities tends to be fragmented, and has not been addressed in a unified way. This
book will make a conscious effort to address the work of the different communities in a unified way.
The book will start off with an overview of the basic methods in data classification, and then discuss
progressively more refined and complex methods for data classification. Special attention will also
be paid to more recent problem domains such as graphs and social networks.
The chapters in the book will be divided into three types:
• Method Chapters: These chapters discuss the key techniques that are commonly used for
classification, such as probabilistic methods, decision trees, rule-based methods, instancebased methods, SVM techniques, and neural networks.
• Domain Chapters: These chapters discuss the specific methods used for different domains
of data such as text data, multimedia data, time-series data, discrete sequence data, network
data, and uncertain data. Many of these chapters can also be considered application chapters, because they explore the specific characteristics of the problem in a particular domain.
Dedicated chapters are also devoted to large data sets and data streams, because of the recent
importance of the big data paradigm.
• Variations and Insights: These chapters discuss the key variations on the classification process such as classification ensembles, rare-class learning, distance function learning, active
learning, and visual learning. Many variations such as transfer learning and semi-supervised
learning use side-information in order to enhance the classification results. A separate chapter
is also devoted to evaluation aspects of classifiers.
This book is designed to be comprehensive in its coverage of the entire area of classification, and it
is hoped that it will serve as a knowledgeable compendium to students and researchers.