No comments about An Introduction to Systems Analysis Techniques (2nd Edition).

2 comments about Modern Control Design With MATLAB and SIMULINK.

1. I loved the book.
   Great explanations.
   Not a complete first book in controls though. It helps to have a little more background or have read one of the standard control's texts.



2. I found this book by chance in the library in our company. I'd studied aerospace engineering and specialised in flight control about 7 years earlier, and this book was perfect for me. It provided a refresher of the basics, and was a good introduction to the more advanced concepts like LQR, H-infinity control, etc., as well as helping me dust off my Matlab skills. But it might not be the best book for newcomers to the subject - but it does complement other books (like Ogata, Stevens and Lewis) very well.

1 comments about Global Sensitivity Analysis: The Primer.

1. When I was working at Oak Ridge National Lab in the late 1970s, I worked with other scientists and statisticians on data and model validation. Toby Mitchell a specialist in experimental design was developing sampling techniques to use in model validation. He has since passed on. One of the techniques he used was Latin hypercube sampling. The authors of this text are from Italy and Canada. They have computer science and mathematical backgrounds but are not statisticians. Yet once you start reading book you will see that they appreciate both the deterministic and the stochastic aspects of modeling.
   
   What they do, they call Global Sensitivity Analysis. They are very bright and are lucid in their explanations and description of philosophical issues. This is not something that those of us who do statistical modeling are very familiar with but it is important to know. It is especially gratifying to see that these authors are always wary of modeling assumptions and look for novel ways to test them. They point out that validating models is complicated. Many models that we construct are complex and even when we are aware of this and test aspects of the model. we often take some things for granted and accept aspects of the model as a given. I really enjoyed reading the afterword where these issues are well brought out.
   
   Specific methods include th elementary effects method discussed in Chapter 3 that is based on the work of Max Morris in a 1991 paper in Technometrics. In Chapter 4, they cover variance-based methods which relies on the work of Sobol and others. They illustrate the applications of these methods with an infection dynamics model.
   
   The main idea is to determine which factors affect the output variables as well as which interaction effects play a role. In the last chapter the authors make recommendations as to when to apply each technique.
   
   At the beginning of the sections they raise questions that they answer in the section. This approach and the problem sets followed by complete and clear solutions makes the text readable and enjoyable even for novices.

4 comments about A Probabilistic Theory of Pattern Recognition (Stochastic Modelling and Applied Probability).

1. This is an awesome book, the best in-depth book on statistical classification to date. Filled with theorems and proofs on classical nonparametric techniques plus neural networks and learning. Standard reference for anybody doing serious pattern recognition and learning. Destined to become a classical reference in the field.



2. This book provides a solid theoretical foundation for pattern recognition and statistical learning. If you consider yourself and expert, or want to be an expert in this field, this book is a must read. It will make you think hard about the concepts (and may be question whether you are or want to become an expert!).



3. The book is great but the notations the authors employ will make you want to drop it on a first reading. Despite the generic title, it is really a reference book for the experts.

   Issues in generalization are presented better in the book by Anthony and Bartlett but overall it is the best book available (for learning theorists).




4. In giving this book a second read, its importance finally dawned on me: it is one of the few if only books that provides a well-rounded theoretical (i.e. mathematical definitions and proofs) perspective on pattern recognition. Although other books, such as Duda et al's "Pattern Classification", have a significant degree of mathematical rigor, very few can claim to be based on the solid mathematical foundations of Lesbesgue measure theory, as this book is. This book has been a big inspiration for me, in that most pr papers I come across provide some method X, and show how experimentally it is more efficient or effective than methods Y and Z. Such papers, although possibly generating interest in the subject or method, do little if anything to advance the theory which in the end will have the final say of how, when, and why something works or when it doesn't. On the other hand, by making the assumption that the data comes from an unknown (i.e. nonparametric) probability measure space which induces an inherent optimal Bayesian error on the classification problem, this book shows how the theory of probability can be used to prove some very interesting results.

   As an example, the authors define what it means to have a universally consistent classifier; i.e. a classifier which converges to the optimal Bayesian classifier as the amount of training data approaches infinity in the limit (irregardless of the data distribution). Moreover, one of the important results is that such classifiers exist and are often quite easy to devise (e.g. nearest-neighbor methods). And to be able to mathematically prove this is indeed inspiring.

   In closing, I would highly recommend this book to anyone who has the mathematical prereqs (probability from an abstract measure-theory point of view)
   and is interested in doing high quality mathematical research in pattern recognition. For that audience this book will provide a good foundation for literally an unlimited number of interesting questions; many of which remain unanswered.

   For those who are more interested in the practice of pattern recogition, the above mentioned book by Duda et al. (ISBN 0471056693) will do just fine as a reference. The book "Pattern Recognition" by Theodoridis et al. is also of high quality (ISBN: 0126858756).

5 comments about Signals and Systems (Prentice-Hall Signal Processing Series).

1. The book has been arrived in good condition and spent a time less than I expect.Thanks for all.



2. Amazon's editorial reviews are correct. Don't be deceived into thinking that this is anything other than an introduction for the complete novice. It's a book with a particular style: long and wordy. It's only for a particular *kind* of novice: one who needs a lot of hand holding and every detail worked out. It's nearly 1000 pages. I learned the subject in 1976 from an earlier edition that was a third the size (I think) but there doesn't seem to be twice again as many new topics covered.
   
   The mathematics is at an undergrad level, with much (but not all) developed as needed in the text. Topics are missing. For example the inverse Laplace transform is mentioned but not developed (perhaps an appropriate choice in a book for the novice), and the Butterworth and elliptic filters are mentioned, even with graphs of their frequency responses, but they are not defined in any way whatsoever.
   
   Some topics are reserved for the exercises. For example, windowing is covered only in one long-ish problem.
   
   The Amazon review says it's a good book for self-study. I'd agree, provided you understand that you are getting only an introduction at the undergraduate level. If you are a grad student or professional, or if you can't tolerate long detailed explanations, this is not the book for you.



3. This book explain very clear what are differences about Fourier and Laplace Transform. Eventhough, those transform appears in the same range of time.
   The writer of these book is well known Professors who is also editor of Prentice hall series in Signal processing.Eventhough, I just past these subject two times in Bachelor and Master coursework. I don't found it is useful for my research until recently. I should do more exercise in this book!.



4. I have struggled through two classes utilizing this book, so I feel I do have a good sense of the material in it. It does say something about a textbook when your professor requires you to purchase a second book to fill in the gaps this book creates (Laplace transform). I feel I still do not have an adequate grasp of Fourier series - and I've read those sections time and time again. I eventually went to another textbook to teach myself the material on my own time.
   The lack of mathematical examples make the processes of understanding the material WAY more difficult than it needs to be. Another sore point: A lot of the understanding comes from the practice problems. Too bad there are so few answers in the back of the book. It leaves you to question what you have learned - if anything from the problem.



5. I am an Aerospace graduate student reading this book on my own time and pace (not taking the undergrad Electrical Engineering class that teaches/uses this book). I found out about this book because my friend (an EE student) was in that class and I asked him if I could look through it. After looking through it I decided to buy it, and I'm glad I did.
   
   This is a VERY GOOD no-nonsense book. A brief personal background, I've read Control Systems Engineering by Norman S. Nise (Very good book!), and am now reading this book. I would say that this book is really not something you want to dive into without any prior background. Start with a good controls book (Nise, Ogata, etc) and learn the basics. THEN read this book to get the finer details.
   
   Also, make sure you have the MATH background for this book. A lot of this book is dedicated to the fourier series, laplace xfrm, and z xfrm. If you have not had a formal class is fourier series, you might find this difficult to grasp. I would STRONGLY recommend you have a background of: SISO controls, ODEs, and PDEs.
   
   If you have already taken these pre-reqs, this book is a great 'aha!' moment that nicely combines the concepts from all these areas and really gives you nice insights into how they are all related.
   
   As for the problems, I'm not doing them. I'm reading this book for the material, not for a grade. I don't doubt the objections made that the problems are SIGNIFICANTLY harder than the examples. BUT, life is tough. I would recommend you look through my PDE book (Partial Differential Equations - Strauss) to see the BIG JUMP in the hw from the VERY skimpy examples. There are lots of books like this. They are made to be hard for a reason, you have to use your brain to think creatively. Dont expect every book to hold your hand through problem solving!
   
   I would ignore the people who rate this book a 1 just because they cant solve the homework problems. The writing is VERY clear and to the point. Please note, this is NOT a book you want to rush through when reading. If you take your time you will find that its actually VERY insightful.
   
   Given the fast pace of a normal semester, you might find this book to be very dense. In other words, you will have to absorb a TON of material in a short amount of time. So I can see this being a tough class to take. Reading it at your own pace, this book is pure bliss!

5 comments about Creating Computer Simulation Systems: An Introduction to the High Level Architecture.

1. Agree with the reviewer about "pompous computer speak", and that the real info could be presented in about 10 pages. I have far too many docs that I need to wade through. The published IEEE standards are always available if you're having insomnia. BUT, with some motivation to get through this book, I was able to accomplish my objective: get a feel for the HLA. [...] Next I'll try the Singhal/Zyda book that another reviewer suggested.



2. I am at the mid-point of the book and will likely post another review when I am done. [...] The authors give the history and motivation and design decisions behind HLA. They also give many good examples, [...], allow you to get a really good feel for the important concepts of HLA by running an actual federation. One thing I will be looking for is the impact of the architecture on simulation performance, scalability in practice (as opposed to in theory), and how is HLA likely to evolve over the next couple decades.
   I don't know yet whether the book is enough for you to create your first federation. If you really have NO background at all in simulation, you will still get a lot out of the first couple of chapters, plus the many references to articles written on the subject, but don't expect to find the other chapters easy. Using my background in simulation systems, I can say that HLA seems to have been very well thought out, based on real-life simulation systems, and is therefore not trivial. But that's what makes it interesting, and the book so far lives up to that.



3. This book covers a lot of the basics of what HLA is but contains holes. After reading this book three times I am still left with questions as to how to implement HLA in an application. I [...] still have unanswered questions as to how to implement the HLA. The examples in the book are implemented in Java [...] however I would have been happier if the examples were implemented in C++ as well. This is a good starting point, in the absence of alternatives, but could be better.



4. To paraphrase Douglas Adams, it is very easy to be blinded to the essential uselessness of the HLA by the sense of achievement you get from getting the RTI to work at all. In other words -- and this is the rock solid principal on which the whole of the DMSO's DoD-wide success is founded -- their fundamental design flaws are hidden by their superficial design flaws. This book covers in detail all of the HLA's superficial design flaws while omitting coverage of the fundamental design flaws. Thus, give it a miss.



5. I think most of the negative reviews are actually against HLA and due to a lack of experience with component-based simulation system development ... not the book itself.
   
   With the lack of resources out there, I really appreciate this resource.
   
   I've just begun to read the book and find it very easily read. I'll post another review as I progress.

5 comments about Dynamics Of Complex Systems (Studies in Nonlinearity).

1. This book is designed as a text to introduce graduate students in science to the concepts and methods in the ``science of complexity'' which comprises studies in mathematics, physics, chemistry, biology, computer science, sociology, psychology, economics, anthropology, and philosophy. Written from the perspectives of a physicist, definitions are informal; thus a concise definition of a complex system is not given. The concept of a complex system is introduced through examples, and informally described as having ``a large number of interacting parts'' although ``even a few interacting objects can behave in complex ways.'' More precisely, complexity is defined as ``the amount of information necessary to describe a system.'' Another key concept is the phenomenon of emergence which arises when ``the collective behavior [of a complex system] is not readily understood from the behavior of its parts.''
   
   Dynamics of Complex Systems opens with a long chapter (278 pages) of ``introduction and preliminaries'' which surveys iterative maps; thermodynamics and statistical mechanics; activated processes (glasses); cellular automata; statistical fields; computer simulations; information theory; computation; and fractals, scaling and renormalization. It is suggested that this chapter can serve as the basis for a one-semester course. This introductory chapter is followed by eight chapters devoted two each to four different subjects: neural networks, protein folding, biological evolution, and human civilization. In each of these pairs of chapters, the first is more detailed and the second more general. Thus the first of the two chapters on neural networks describes neural network models (Hopfield's attactor models) whereas the second discusses the phenomenon of sleep and models of mind, with similar divisions of labor in the pairs of chapters on protein folding and on biological evolution. In the final chapter, it is noted that ``human civilization is more complex than we are as individuals.''
   
   Alwyn Scott
   http://personal.riverusers.com/~rover/



2. That physical systems are complex has been acknowledged for centuries, but only in recent decades has the scientific community, especially physicists and biologists, directly confronted complexity. This book discusses complex systems from the dynamical systems perspective, and as such can be read by physicists, mathematicians, and mathematical biologists. Biologists in particular will find the discussion of `emergence' the most important one, especially systems biologists. Physicists and mathematicians who study dynamical systems tend to not be concerned with their origins, whether they are in biology or some other area. But physicists do concern themselves with the experimental relevance of dynamical systems, unlike mathematicians who are sorely concerned with their formal properties, and do not care at all if they can find expression in the real world. But it goes without saying that the theory of complex systems has found application in finance, genetic engineering, cryptography, network engineering, and many other areas. This book gives a good overview of the techniques used to study complex systems, and can be read by anyone with the necessary mathematical preparation, consisting of probability theory and elementary calculus.
   
   Systems that are simple can become complex by only a slight alteration in their configuration. The gravitational three-body system in classical mechanics is a good example of this. The dynamics of two objects interacting gravitationally can be solved explicitly, but the system consisting of three bodies cannot. The complexity in these two cases is measured by the availability of solutions to the dynamics of the system. The author is very aware that more involved measures of complexity are needed and he gives examples of these in the book. Mathematical techniques from probability and statistics are of course used throughout the book to frame these measures more quantitatively. This reflects the author's stated strategy throughout the book, namely to describe the essential characteristics of a class of systems, and employ statistical techniques to find the properties and behaviors of these systems.
   
   The concepts of emergence and complexity are fundamental to a study of complex systems, the author argues and early on in the book he clears up some of the confusions behind the use of these terms in the scientific literature. A `complex system' is one which is constructed from many components and whose behavior cannot be determined from the behavior of these components, i.e. the behavior of the system is `emergent.' The `complexity' of a system, on the other hand, is the amount of information needed to describe the system. This is a somewhat subtle definition, and quite a few proposals have been put forward in the literature for measuring complexity. The author settles on a familiar method, the `entropy' for measuring complexity, but with a warning to the reader that the calculation of the entropy is dependent on the particular length or time scale over which the system is observed. For extremely long time scales (of observation), one can get away with describing systems as always in equilibrium. In this case the entropy would be maximum but the system would not be viewed as being complex. For very short time scales (of observation) , the entropy of the system is very small but due to the ability to observe the microscopic dynamics of the system it would be viewed as highly complex.
   
   These considerations lead the author to introduce the concept of a 'complexity profile' of a system, which he discusses at some length in the last pages of the book. The complexity profile is designed to study the the dependence of complexity on both length and time scales. The concept is dependent on the notion of a sequence of observers that are ordered according to their ability to distinguish microstates. The author calculates the complexity profile of the ideal gas and shows that the complexity of a microstate for this case is simply the entropy, but as the number of microstates with a given region increases, the complexity approaches zero. Other examples of the complexity profile are discussed, one being for observers that only measure the positions of particles and not the momentum. The author also studies the connection between the complexity profile and the predictability or chaotic behavior of the system, where chaotic systems are viewed as being ones where information from a particular scale can be transferred to a larger scale, as contrasted with dissipative systems where information on a large scale is transferred to a smaller scale. The author gives various arguments and calculations that illustrate the difference in complexity profiles between chaotic systems and those of conservative, nonchaotic systems. The discussion is fairly convincing but if the complexity profile is important in complex systems, its defintion and properties should have been included at the beginning of the book, and serve as a central theme behind the discussions throughout the entire book. As it stands the complexity profile comes across as a concept that is purely ancillary to the study of complex systems. It certainly does not appear to be indispensable in discussing irreversibility of physical systems, this problem still being the most pressing one in statistical mechanics and is still hotly debated at the present time.



3. that information is the opposite of entropy which is a measure of disorder or uncertainty. However because this book is about complexity and not information per se, I will only briefly refer to his mistakes with the latter as I have explained them further in other reviews that are specifically on that topic.
   
   Shannon's information rate from communications theory, R, is an entropy like formula but most critically it is a state function difference of the uncertainty reduction to a recognizer after a measurement. Entropy is not a proper measure of disorder or uncertainty; the 2nd law of entropy increase of the universe applied long before there were any observers. It is a measure of the dispersal of energy. Going back in time is not going back to perfect order, but quite the opposite. I have not seen proper definitions in any book but there are PhD level articles available on the internet with proper definitions such as the Principia Cybernetica Web and molecular biologist Dr Thomas Schneider's website. Biologist Richard Dawkins also has an accurate short article on the internet. Most physicists have the definitions wrong unfortunately and believe information evolved before life, which is false. (A recognizer is required, whether a ribosome or mind etc.) Instead a better definition of complexity than the present author offers would indicate that the universe has increased in complexity through gravitational clumping (among other things). By making the mistake then the physicists and present author believe maximum information is randomness or equilibrium. This is the definition of algorithmic complexity.
   
   As the author adapts algorithmic theory to his complexity profile he arrives at formulas that are observer dependant: "the complexity profile [is] the length of the description [of] the error allowed [as] the description increases." This is of little or no practical use. Again the universe has grown in complexity (or at least in pockets or we wouldn't be here) without relying on the degree of focus of any observer. A crystal is highly ordered relative to say a human cell whose complexity is a result of a multitude of interactions of chemical agents and macro molecules. This is where his analysis falls silent, in fact wrong. He says (page 741) "short-range correlations decrease the microstate complexity..." Well that's because he has a flawed method of using statistical mechanics. There is likely no universal complexity algorithm. Consider that a single gene can yield up to thousands of different proteins. One should be wary however of any formula that treats correlations as reduced complexity! Again the crystal vs the cell!
   
   However there are ways of measuring the critical biological requirement of interactions that in fact increase complexity, the opposite of equilibrium statistical mechanics, a flawed tool. For instance in a recent article at lanl.arXiv.org, authors Edwin Wang et. al. apply Pearson's correlation coefficient to show that "genes with higher cis-regulation complexity are more coordinately regulated by transcription factors at the transcriptional level and by micro RNA's at the post-transcriptional level. This is a potentially novel discovery of a mechanism for coordinated regulation of gene expression...We found a positive correlation between these two groups of transcriptional regulators... " Measures of correlations are key in studying biological complexity and are not based on an observer's focus ability.
   
   For a layman's guide to the issue of correlations for life see Irun Cohen's book 'Tending Adam's Garden' (though it has no quantitative aspect).



4. The book is a tour around the paradigms used by scientists in
   Complex Systems. While normal science is about using and re-using the paradigms without much creativity or true aportation to knowledge or understanding, the situation is worse in complex systems, since, as an emerging area it has multiple (competing?) paradigms, to the point that it is not possible to define a "complex system" in a form that encompases all the paradigms. The book certainly does not solve this problem, yet the author acknowledges the difficulty present in saying what a complex system is.
   Complex systems is not an area of research but a community of researchers united by their interests.
   The book is then a compilation of the "how to" and the believes for each paradigm, several of them carry very little science and have left the idea of "refutation" burried under piles of meaningless papers. Not surprissingly, some authors claim that complex systems is a postmodern scienceComplexity and Postmodernism: Understanding Complex Systems. And truly, the complex systems of Bar-Yam are only possible after we have buried reason and have accepted that science has nothing to do with truth.
   Too much for me, not a book I recommend to my students.



5. This is a beautifully written and thought-provoking work that presents the field of complex systems in a unified manner. The writing is highly engaging and stimulating with a broad range of topics. The material is pitched at just the right level, focusing on the concepts without getting buried in unnecessary details, while avoiding superficiality. I highly recommend this excellent book.

5 comments about Complexity and Postmodernism: Understanding Complex Systems.

1. The book combines elements of different philosophies: post-modernism, structuralism, and deconstruction. It is a meeting of vague philosophical generalizations and scientific terminology (e.g., neural networks), and as such, it muddles things instead of making them clear. The hope being that, if things look complex and muddled, people will consider the book profound.

   I have to say that stylistically the book is fairly well written, yet this is not something one would read for entertainment. Bottom line: this is an attempt at some sort of philsophical synthesis which, in reality, is an intellectual dead end.




2. I read this book primarily through an interest in the philosophy of language. Of particular relevance in this respect is the emphasis on a characterisation of complexity as being opposed to traditional notions of representation. Cilliers draws parallels between the philosophy of Saussure and Derrida and scientific developments in distributed representation, particularly with respect to connectionist approaches as implemented in neural networks. Cilliers argues that a classical representational theory of language that posits syntax as an instantiation of semantics does not sufficiently allow for the complexity evident in language, but rather that meaning is constituted by the dynamic relationships between both the components of language and the environment in which it is embedded. Cilliers explicitly rejects rule-based symbol systems as being adequete for modelling language, referring to recent scientific research using neural networks to simulate language learning indicating that "though rules may be useful to describe linguistic phenomena, explicit rules need not be employed when language is acquired or when it is used" (p. 32). In Chapter 4 (pp. 48-57), Cilliers considers the Chinese Room Gedankenexperiment from the perspective of his thesis. He suggests that the debate has unquestionably assumed that the formal model of language represented by the argument is correct, that is, that a rule-book such as the one supposed is even possible. Cilliers suggests that this assumes certain features of language: that a formal grammar for a natural language can be constructed and represented in a lookup table; that there is a clean split between syntax and semantics; and that language represents rather than constitutes meaning (p. 53).

   The overall picture of language that Cilliers develops has important parallels with the views of Wittgenstein, though, somewhat surprisingly, Wittgenstein is never explicitly mentioned (except with regard to his family concepts). Firstly, meaning is construed as occuring through dynamic processes (use) rather than static representations (the conception that Wittgenstein's private language argument criticises). Secondly, the idea that there is some fact of the matter (whether inside or outside human agents) that determines meaning is explicitly rejected. Finally, a straightforward split between syntax and semantics is denied (a distinction that the sceptical interpretation of Wittgenstein, offered by Kripke, takes advantage of).

   In summary, I would recommend this book to anyone interested in making connections between dynamic systems theory and philosophy of mind or language -- Cilliers proves an effective communicator in both of the fields he wishes to connect.




3. Frankly, I'm astonished by some of the favorable reviews this book has received. First of all, I still haven't figured out if this really is a book or if it's a collection of essays, due to the amount of repetition of content between chapters.

   Cilliers attempts to demonstrate the mutual relevance of complexity science (CS) and postmodern philosophy, but his knowledge of CS and thermodynamics seems to go no deeper than what he's read on the dustjackets of pop-sci books. The number of claims he makes that are either blatantly false or not necessarily true are outnumbered only by the number of uninsightful comments and statements that appear to have been gleaned directly from more technical sources. Here are a few to make one's skin crawl:

   On p. 6, as an example of a non-linear relationship: "money can receive compounded interest". In fact, this is a classic *linear* relationship (so common it's often used as an introductory problem the first day of a course in linear differential equations). The equation representing it is simply: dM/dt = n*M, where M is the amount of money in an account, and n is the interest rate. The solution is Mo * e^(nt), where Mo is the initial amount of money in the account and 'e' represents 'exponential'. (Simply because compounded interest generates an exponential curve over time does not make the relationship non-linear; the underlying equation is linear.)

   On p. 4: "Any analysis of a complex system that ignores the dimension of time is incomplete, or at most a synchronic snapshot of a diachronic process." This is completely false - One of the very purposes of 'phase space' analysis is to *completely* represent a system without considering time. The elliptical relationship between velocity and momentum in a simple harmonic oscillator is a common example that many might remember from high school physics.

   On p. 8: "In classical mechanics, time was reversible, and therefore not part of the equation. In thermodynamics time plays a vital role." This quote still makes me tear at my hair. The *exact opposite* is true: almost every equation in classical mechanics (projectile motion, harmonic oscillation, planetary motion) explicitly involve time as a dimension, while, because thermodynamics is only concerned with initial and final (equilibrium) states, few thermo equations do so.

   On p. 3, Cilliers says: "The grains of sand on a beach do not interest us as a complex system." but includes later in the book a quote from complexity scientist Per Bak, who has achieved his fame specifically for the study of the 'self-organized criticality' of sand grains.

   And this is just the first few pages! The list goes on and on: He repeatedly confuses the thermodynamic concepts of 'closed' and 'isolated' systems; He seems to think that 'non-linear' equations are all somehow phenomenally complex and unsolvable and that the phrase 'non-linear' is therefore a synonym for being non-reductionist, non-rational, and, in short, 'postmodern'. (In doing so, he falls into many of the traps Alan Sokal identified in Fashionable Nonsense.)

   I think that the basic concept behind the book could have been interesting, but due to Cilliers elementary-level grasp of half the subject matter with which he deals, the statement Cilliers himself makes on p. 133 (in reference to a recent book by Rouse) applies equally well to this text: "For me, reading this book was about as pleasant as it would be to eat it."




4. ... in spite of the appearance of the reviews associated with this work and the work itself, there is a valid connection between postmodernism and (let me be patient!) complexity.

   First of all, about terminology... isn't complexity theory a branch of computer science that deals with execution time as a metric of algorithms? I think the reviewers here want to refer to complex systems theory. Wasn't connectionism a fad which was piled on top of a catchily-conceived name for artificial neural networks .... which were the popularization of more serious works of people like Papert, Minsky, Grossberg...and doesn't the reviewer who pretends to know something about physical science understand what "irreversibility" is and that, indeed, classical mechanics is indeed reversible? J. Willard Gibbs would roll over in his grave if he could read the reviews on this page...

   IF you are seriously trying to find out what this stuff is about, start out by getting Lars Skyttner's book on General Systems Theory. Use it as a guidebook. Then, if you want to understand the evolution of the ideas, read the opening sections of Kant's Critique of Pure Reason. After that, read Saussure and Piaget on structuralism and read Terence Hawkes' book, "Structuralism and Semiotics" After that, try to get at least a rudimentary understanding of the work of the process philosophers...Bergson, Peirce, James and, of course, "Process and Reality" by Whitehead. At this point, you should seriously consider getting at least a passing familiarity with the work of Karl Marx with the goal of understanding what was really bothering him - and of seeing that Marx's ideas are important in ways that he probably never even thought about.

   At that point, if you are one of many for whom there is a schism between the culture of liberal arts and the culture of mathematics and science, you should, at this point read a few of the popular works of Richard Feynman - perhaps, "The Character of Physical Law" or the opening lecture of Volume I of "The Feynman Lectures on Physics". Compare what Feynman has to say about science to what Piaget has to say about structures and - hopefully, by now you are beginning to realize that mathematics is a liberal art - and that the so-called liberal arts are sometimes excuses for people who don't want to be very careful in their thinking....(not always, mind you) - go and read Sunny Auyang's wonderful books, "Foundations of Complex Systems Theories" and "How is Quantum Field Theory Possible?"

   By this point, you should feel somewhat secure in addressing "Postmodernism" and being able to distinguish what is there because people want to sound "cool" for their friends, and what is valid and sometimes deeply disturbing for the evolution of humanity.




5. Cilliers does an excellent job to bring together difficult concepts. One or two reviewers come up with issues around time and thermodynamics: let's just say that entropy (central to the second law)is a temporal concept and leave it at that. I found his work something to go back to regularly, well referenced and an example of how to fuse philosophy and science in a rather rigorous manner. We have moved ahead in neural network technology and thinking since 1998, but I still use the concepts of Cilliers to validate thinking around the modern concepts of non-linear estimation and pattern matching. I believe that this is a very good book to use as part of core reading material if you come from a scientific or a social sciences background and need to work on real world problems at that uncomfortable cross-over point between the two.
   
   Why not 5 stars? This is 2006 and we have internalised many of the ideas that Cilliers had to expose and defend in 1998. So Paul, what about a new work that brings in your ideas on "slowness" and bounding of systems?

1 comments about Physiological Control Systems: Analysis, Simulation, and Estimation (IEEE Press Series on Biomedical Engineering).

1. I have been using this text for a class in Physiological Control Systems, but have been largely disappointed. One of my disappointing experiences is on p. 170-1, where Khoo shows how to get an RLC model transfer function out of MATLAB's ss(). Since the MATLAB documentation on ss() is skimpy, this is a place where Khoo could have added value, illuminating what the A, B, C, and D matrices represent to ss(), but Khoo simply brushes past the opportunity. Khoo also discusses bifurcation in the logistic map, but if you look for 'logistic' in the index, you won't find it. Khoo mentions Fitzhugh-Nagumo and Hodgkin-Huxley within the context of his section on Bonhoeffer-van der Pol, but those four authors are not in the index (Bonhoeffer and van der Pol are). I admit to not having made a comprehensive study of the MATLAB examples, but I downloaded his code for sensitivity analysis (sensanl.m and two supporting .m files) mentioned in section 7.3.2, and consider the code to be poorly written. If I didn't have Dorf & Bishop's "Modern Control Systems, 9th Edition" to fall back on, I would have been in dire straights getting anything beyond a cursory reading out of Khoo's text. In short, this book should command a price in the $50 to $60 range, not the stellar $110-120 its priced at. Dorf & Bishop is priced about the same and delivers three times the value that Khoo does. Every chapter where I made an effort to get to the bottom of some discussion, I found Khoo's exposition wanting. The index is exasperatingly useless. There are only two entries under 'H', one under 'K', one under 'W', etc. That's alarming for a book with 307 pages.

4 comments about Global Positioning Systems, Inertial Navigation, and Integration.

1. Of course the experience of the authors is great but none of it you can use either in your study or research. It is impossible to learn about Kalman filtering from this book as well as to improve your knowledge in this area. Book is intended for professionals in this area but professionals will not need it. Many intersting facts but no detailed information. For example one can know that the square root filtering method came from James Potter but there is nothing more about this method from this book. Otherwise you have to study sources of the software (included). Book is written as advertizing of author's skills and nothing else. Too much about nothing ...



2. Here are some quotes from a review of this book in GPS World, July 2001, pp 46-47, by Dr. John Angus, a consultant and researcher in the area of GPS-aided navigation systems, AND professor at Claremont Graduate University, Claremont, CA. Dr. Angus' credentials qualify him to review this book.

   "Noteworthy is the comprehensiveness of the material on GPS, Kalman filtering and Kalman filter engineering, and the appendix on coordinate transforms...An instructor could easily develop a one-semester course on basic GPS or a full year course on GPS and inertial navigation, each of them "glued" together by the Kalman filter and enlivened by computer experiments with the MATLAB code provided."

   "The writing...tends to be concise and the mathematics is kept to the minimum necessary to expose the theory and methods of filtering, GPS, and INS."

   "...effectively addresses most of the basic engineering and performance issues relating to GPS/INS."

   "...recommended for personal and professional libraries."

   This is an application-oriented book, which as such, does not include detailed mathematical derivations. It does provide Kalman filter algorithms (on floppy and in text), but if one needs the theory of Kalman filtering behind these, one needs to use a Kalman filtering text, such as Kalman Filtering Theory & Practice Using MATLAB (Second Edition), Wiley 2001, by Grewal and Andrews. The latter book gives all of the methods in square root filtering algorithms and derivations and more. If the "Asian Reviewer" is most interested in Kalman filtering, he/she would be better advised to buy a book on Kalman filtering.




3. Here are some quotes from a review of this book in GPS World, July 2001, pp 46-47, by Dr. John Angus, a consultant and researcher in the area of GPS-aided navigation systems, AND professor at Claremont Graduate University, Claremont, CA. Dr. Angus' credentials qualify him to review this book.

   "Noteworthy is the comprehensiveness of the material on GPS, Kalman filtering and Kalman filter engineering, and the appendix on coordinate transforms...An instructor could easily develop a one-semester course on basic GPS or a full year course on GPS and inertial navigation, each of them "glued" together by the Kalman filter and enlivened by computer experiments with the MATLAB code provided."

   "The writing...tends to be concise and the mathematics is kept to the minimum necessary to expose the theory and methods of filtering, GPS, and INS."

   "...effectively addresses most of the basic engineering and performance issues relating to GPS/INS."

   "...recommended for personal and professional libraries."

   This is an application-oriented book, which as such, does not include detailed mathematical derivations. It does provide Kalman filter algorithms (on floppy and in text), but if one needs the theory of Kalman filtering behind these, one needs to use a Kalman filtering text, such as Kalman Filtering Theory & Practice Using MATLAB (Second Edition), Wiley 2001, by Grewal and Andrews. The latter book gives all of the methods in square root filtering algorithms and derivations and more. If the "Asian Reviewer" is most interested in Kalman filtering, he/she would be better advised to buy a book on Kalman filtering.




4. One of the students in my lecture came up to me and said he had studied this book all summer to learn about GPS. He found it to be very helpful to him and he said it was the best source of information on GPS he had come across.