Customer Reviews

Bioinformatics: The Machine Learning Approach, Second Edition (Adaptive Computation and Machine Learning)

5 star:		(10)
4 star:		(0)
3 star:		(1)
2 star:		(1)
1 star:		(4)

 

 

I just bought this book and am COMPLETEly disappointed with it.
Here is why. The book is badly written, hard to read and follow. Although it is said that this is a book is for " many readers", it is really for those who have already known all the algorithms. It is simply impossible to learn the algorithms from this book. The chapter on neural network is a few pages. It provieds a few equations for backpropagation. That is it! It is pretty much true for every thing else. Equations, hard to understand sentences, abbreviations with no explnantions, tons of citations everywhere. A book should strive to explain, and not to cite what other papers and go look there all the time. I suspect the few good reviews here are from the authors themselves.

I have a good programming background. I also read some papers on neural network and hidden markov models, This book is a lot worse than anything I have read in explaining the stuff. Very disappointed. Save your money and get something else.

This book is decidedly a mix: some very good information, combined with some very puzzling omissions and uneven editing.

First, the good. The description of stochastic context free grammars is the best I've seen. I don't know any other reference that even hint at how to use generative grammars to evaluate likelihoods. Once they caught my interest, though, the authors did not carry through with training and evaluation algorithms I could really use. I suspect that parts of the information are there, but I'll have to go back over their opaque notation again to work out just what they've given and just what's been left out.

This same pattern - an interesting introduction with missing or mysterious development - recurs throughout the book. The discussion on clustering and phylogeny goes the same way: a number of techniques are mentioned but not developed. The authors mention a tree drawing problem, not just building the tree's topology, but ordering the branches for the most informative rendering. Again, a critical topic and one that most authors miss - in the end, these authors miss it, too, by mentioning but not filling in the idea.

Their discussion of neural nets suffers badly from the authors' partial presentation. Evaluation of network output for a given input is relatively straightforward, and they present it in some detail. Training the net is the real problem, though, and is given less than a page.

Baldi and Brunak give more of the fundamentals than most authors. For example, they explain the maximum entropy principle well enough that I'll use it in lots of other areas. They give some coverage to topics of intermediate complexity, such as the forward and backward algorithms for HMM training. Finally, they fizzle out at the higher levels of complexity - the Baum-Welch algorithm could have followed from the forward and backward methods, but is left as a reference to another book.

There is some good here, especially in the fundamentals behind important techniques. The discussions I wanted - the more avanced topics, in forms I can use - are often weak, missing, or impenetrable. Just a bit more work, clearly within the authors' capability, would have made this a landmark reference.

I'm a graduate student, reading a lot of bioinformatics materials. This is by far the worst text I've read on the subject. Poorly explained, poorly edited. Poor.

Just a collection of formulae, in an unclear way. Once we tried to use it in our seminar of bioinformatics, but after a few chapters we had to give it up for its bad writing. I could not find any reason to buy it or read it.

This book is abysmally written. It appears to be filled with technically accurate information, but not organized in a form amenable to learning. It is probably not even appropriate for an expert: an expert could probably verify its correctness, but it's not organized appropriately to be used as reference.

I have a PhD in computer science and I'm used to working hard to master technical material without assistance. However, I found it extremely difficult to fight off the feeling that these authors' goal was to parade what they know without actually sharing it.

There's no need to pay extra money for color photographs of the authors' car license plates and a half-breed cat. Buy Durbin, Eddy, Krogh, Mitchison's well-written book instead.

"Bioinformatics", by Baldi and Brunak, is a very well-written treatment of current stochastic algorithmics of genomics and proteomics. It is profitable reading for both the computer scientist learning relevant biology and the computational biologist learning relevant computer science. It probably favours the biologist slightly in this regard, as witnessed by my own enthusiasm for this work. Of particular value are the chapters on hidden markov processes and stochastic grammars. The treatment builds smoothly from early chapters on Bayesian fundamentals in chapter 2, to markov chain monte carlo processes in chapter 3, followed by theory and applications of neural networks, three chapters on hidden markov processes (a fascinating and vital field in modern genomics) and lastly an introductory chapter to the equally important area of stochastic grammars. Other appreciated features include: an up-to-date 452-reference bibliography; a comprehensive survey of web-based resources re both genomic databases and available search engines for DNA, RNA and protein sequence-patterns; in the appendices, there are concise definitional reviews re the coupling of information theory with entropy and aspects of HMM's.Lastly, the price is right, as is most often the case with books from MIT Press.

The above authors have succeeded well in illuminating a large piece of a very large (and growing) object: the landscape of modern informational biology. They of course cannot cover it all. Another recent book (1997) that complements this book's particular focus is that of Setubal and Meidanis ("Introduction to Computational Molecular Biology"). These authors offer a greater emphasis on string and graph theoretic approaches to sequencing algorithms and deal more directly with various heuristic approaches to fragment assembly and hybridization mapping.

I found the book very readable, and full of information combining the machine learning approach (neural nets and Hidden Markov models) with biological problems. The wealth of specific biological information bridges the background gap for computer scientists and mathematicians, and the organization of topics is excellent.

In the mathematics and computer science community, Baldi is an internationally recognized expert in the fields of neural nets and Hidden Markov models and their applications (for instance, he holds a patent for neural net recognition of fingerprints). Concerning HMM's Baldi and co-workers have found statistical models for protein families, sequence signals for nucleosome centers, etc. Moreover, Baldi, together with Chauvin, has developed a gradient descent parameter update method for HMM's which has no zero probability absorptions, and allows on-line updates, useful features not supported by the standard EM method.

From these and other applications, I found the text very useful.

This book is an excellent source of information for beginning the study of machine learning algorithms applied to biology. Reading the book you get a clear feeling that bioinformatics is not just one of the many application fields of computer science and artificial intelligence, it is perhaps the most challenging set of problems for intelligent algorithms not primarily focused on replicating human intelligence. There is an amazing wealth of open problems, some of which apparently very difficult. No doubt that unless you are already an expert you need an accurate map of this complex territory and the book by Baldi and Brunak is an excellent and up-to-date map that may suggest new exciting ideas for research.

As a computer scientist I can say that the book is sometimes difficult to read if you have no previous knowledge of biology. This is because the authors didn't take the simplificative approach of reducing biological problems to abstract mathematics. Rather, they preserved the full biological flavor of the problems. Although this approach costs you more at the beginning, you can eventually get a more accurate and nontrivial picture of the problems.

My conclusion: it is perhaps unlikely that you can learn about bioinformatics using only this book. However, if you want to learn about bioinformatics, this book is a must-have reference.

The book provides an abundance of excellent information of machine learning techniques as applied to biology. I found the presentation of the material to be clear, detailed, with a wealth of support data regarding many of the complex issues of BI. Thanks to Baldi and Brunak, the ideas such as hidden Markov models and applications in molecular biology are dramatically clear.

This is an excellent book. It contains a broad introduction to the main problems of computational molecular biology and a rigorous description of the foundations of machine learning and other statistical methods. Several chapters cover a variety of applications from DNA, to RNA, to proteins problems. Unlike other books on these topics, this book has very few errors in it, if any.