From the Publisher
The book would be enlightening for a statistical reader wishing to learn about the development of more empirical, less formal, methods in parallel to the work being done by the statistical community.
—David J. Hand, International Statistical Review (2011), 79
I very strongly recommend [this] monumental monograph for the classroom as a graduate text or as a standalone [book] for professionals such as engineers and scientists for their research. Dr. Mitsa present[s] the latest developments of data mining in the time domain with extreme simplicity and elegance while offering in-depth exposure to the principles and applications of temporal data mining.
In my research, I find this knowledge particularly useful for remote sensing applications, specifically, in the detection, classification, characterization and imaging of distant objects as well as for the detection, characterization, monitoring, and staging of early cancer cells with high discrimination potential and low false-alarm rate, while maintaining adequate sensitivity.
Dr. Mitsa’s invaluable expertise and efforts to enlighten the understanding of temporal and spatiotemporal data mining principles, including the latest techniques on temporal pattern discovery, classification, and clustering, have a tremendous impact on a wide array of multidisciplinary areas of science and technology such as biomedicine, defense, business, and industrial applications.
—Dr. George C. Giakos, IEEE Fellow, University of Akron, Ohio, USA
Temporal Data Mining presents a comprehensive overview of the various mathematical and computational aspects of dynamical data processing, from database storage and retrieval to statistical modeling and inference. The first part of the book discusses the key tools and techniques in considerable depth, with a focus on the applicable models and algorithms. Building on this, the second part considers the application to bioinformatics, finance and business computing. The technical depth is appropriate to interest a broad audience, and the text is highly accessible irrespective of the reader’s prior familiarity with the subject. An extensive bibliography is provided on each of the topics covered, which makes this book a valuable reference for both the novice and the established practitioner. The clear, concise and instructive style will make this book particularly attractive to graduate students, researchers and industry professionals.
—Dr. Wasim Q. Malik, Massachusetts Institute of Technology and Harvard Medical School, Cambridge, USA
… how can decision-makers be so data poor in such a (theoretically at least) data-rich economy? Chapter 7 of Theo Mitsa’s book presents the potential for an interesting resolution to this paradox. Her linkage of sophisticated concepts of temporal data mining to practical business issues, such as strategy, forecasting, financial scenario analysis, customer value and retention, operations and logistics management, etc., offers an illuminating approach to organizing and creating sense from overwhelming quantities of random data. Although the algorithms and computations are complex, a reader can learn that there are quantitative approaches to expose additional, possibly critical, insights about virtually any facet of a business. This book further illustrates the growing importance of business analytics and showcases the myriad opportunities available to savvy managers and entrepreneurs to use a system of tools to leverage the value of, and investment in, their data collection and mining efforts.
—Gary Minkoff, Babson MBA, President, Above & Beyond Marketing, Highland Park, New Jersey, USA
As someone who works on signal processing applications in the medical device industry, I found the topic of temporal data mining to be extremely relevant. Our work focuses primarily on time series analysis of evoked potentials. Analysis of these signals is complicated by interfering signals, which although variable, tend to fall into a fairly small number of stereotypical cases. The techniques described in chapter 2 for temporal data similarity calculations and in chapter 3 for temporal data classification have potential application in our work. I found that Temporal Data Mining offered a valuable overview of these fields and gave interesting insight into topics related to gene discovery and bioinformatics. A major strength of the book is the large bibliography, which provides the reader with the tools to dig deeper into topics of interest.
—Dr. Brian Tracey, Signal Processing Project Leader at Neurometrix, Inc.
