[.ca] Weak Links: Stabilizers of Complex Systems from Proteins ... (ISBN 3540311513)

Primer on real-life networks with a red thread:
If you ever needed another good reason to value your grandmother even more, you'll find the answer in "Weak links". Structurally, his book starts with an exposition on network theory and terminology, then the application and discussion of these concepts to real-life complex systems on many scales and applied to many domains (physical, natural, technological, social). His main point is, as the reviewer noted above, that 'weak' links (weak: addition/removal does not statistically affect the average of some metric) stabilize systems. The book has thorough footnotes, one can delve as deep as one would like into the professional papers. In addition, Csermely is an honest scholar - he shows his hands when there is mere speculation (you have to see the book's unique pictograms to appreciate the effects) After pouring through several alternatives, I have adopted this book as a textbook for my Science of Networks class (I'm CS fac at an elite US liberal arts school), and I recommend it to anyone without hesitation for a readable, and learned exposition. I only have two or three caveats from a specialist's point of view: The phenomenological discovery of power laws in complex systems is not unusual and may not be evidence of any SF properties. Scale-free is an abused term, and I wish the controversy about it were explained a bit more. Also, from a modelling point of view, I wish Doyle and Carlson's work on HOT systems were discussed in more depth. But these are minor points, relatively speaking. This is a gem of a book: erudite, humane, funny, accessible and thoroughly fascinating. On every page, there are delights that lead down new intellectual paths. Csermely did a great service to pedagogy and to this budding science with this magisterial survey. Outstanding in its ease of access for intelligent undergraduates and commendable for intellectual honesty - I wish more books (textbooks and otherwise) were written this way. While I can understand the point of view of the previous reviewer, I do not share the rather dismissive characterization of Csermely's book as the product of a 'mad scientist' and (a bit insulting) '\onon-serious\c researcher'. Network analysis is a relatively young field, and as such, hunches, speculative inquiries are part of the consolidation process, and should not be dismissed so casually. Csermely's credentials should also give some pause.

Interesting ideas, some humour, but lacks coherence and details:
I am studying the brain, and as such I often read about neural networks. I thought that the idea behind this book was interesting, so I took a chance. The author first reviews some basic concepts of networks (basic to those who are well-versed in this, but to others this section would be necessary), and then proceeds in later chapters to introduce his ideas of how weak links influence and stabilize networks, using a variety of research domains as evidence. However, while being interesting ideas with some merit, the book is written more from the perspective of a mad scientist than from the perspective of a serious researcher... and one that suffers from the "my new theory explains all" syndrome at that! The author acknowledges this, and actually means this to be humorous, but this can be a hit-or-miss. He also has sections where he goes into wild speculations (to his credit, these are well-marked as such, so readers not interested in wild speculations can easily skip these sections). Occasional rantings with Spike (his inner annoying voice) also comes in as a humorous distraction - but again, a hit-or-miss. This does not take away from the merit of the ideas though, it speaks solely to the manner that the book is written, and some readers may enjoy this more than I have. The main problem that I have is that some of the ideas are not well presented. I was confused about the chaperones in the preface (not a good start), and I find that many of his ideas could be better supported with either more examples from research, or more details about the studies reported. The content should be made more accessible to people without extensive background in the topics covered, especially, as the author likes to point out, researchers from these different fields of research rarely compare notes. Perhaps a good book to use as a fun introduction to networks with diverse applications, just not exactly what I expected when I ordered it.