Modeling and simulation is an area applied to various fields, whether it be medical data, engine processing data, or simply finding optimization solutions to multiple problem domains. Within this area of simulating systems, there is sometimes the issue of finding the data needed to verify the reliability of the artificial models created. Sparse modeling is a technique that focuses on this area of drawing predictive models in scientific applications based on a limited number of measurements. This book raises a very good case for these problems in the contexts of computational biology, image processing, and social network analysis.

The book is a collection of chapters by various researchers and scientists who use sparse modeling techniques in their own research areas. It is well organized, discussing the problems and solutions to data modeling when there are missing datasets. In the literature, most techniques use neural networks or some training mechanisms to predict the missing datasets; here, the sparse modeling technique uses present data to construct learning and reliability evaluation models to construct the models based on complex mathematical calculations.

The book is of particular interest to me, as it goes into details of sparse modeling and the problems faced when setting up experiments to study and gather data for complex problems. There is the fundamental problem of nonexistent data and how to recover from this, and processing large amounts of data has always been a complex simulation issue. The chapter on sequential testing seems to hit the nail on the head in this respect. The book goes into detail on event detection and how innovative mathematical algorithms and curves can help with estimating and finding the missing data to make better approximations in order to build reliable models.

The book is a very complete, going from the basics, methodology, and case studies to testing using probability theory. Unlike most techniques, it requires Bayesian and various probability methods. The book finishes with various topic models in sparse modeling domains.

Overall, as a modeler of various biological and economic systems, I found the book extremely interesting. The mathematics behind sparse data, prediction probabilities, matrix formations, and analysis for event defections and noise is mind-boggling. With regards to sparse modeling, the book summarizes all of these very well; it is a good starting point for PhD students and researchers. The techniques discussed may be useful in many other domains, especially where finding data and creating models on reliable datasets is a huge problem.