Customer Reviews

AI Application Programming (Charles River Media Programming)

5 star:		(13)
4 star:		(2)
3 star:		(2)
2 star:		(2)
1 star:		(0)

 

 

I enjoyed working through this text, but not without some re-visiting of my calculus classes and trigonometry brush-ups.
All in all a very good book, and also a great Graduate level reference for the inner workings of actual Artificial Intelligence algorithms.
If you are well prepared, this book is to the point, and well worth the read. Prepare for a visit to College-level Physics theorems, as many algorithms given require a working knowledge of the advanced principles of the science.

Hope this helps-

Most of other reviewer think highly of this book. I also agree, to a certain extent, that the book's is valuable and fill in the gap between "talks" and "walks".

However, there are two things I have to point out: One, the editing/basic correctness check of this book is kinda terrible. For example, P72 on Particle Swarm Optimization, the 4.2 formula is obviously WRONG and not consistent with the rest of discussion. Also on P74, the position vector calculation is wrong as well: it also seems the author/editor cut & copy two blocks of text.

Second, I don't like is the lack of explaination on certain important notations and equations, which is very important to be at least "self-contained" for such a "cover everything" book. For example, P210 on reinforcement learning, Equation 9.2 has a general explaination of what it is, but non of those notation/symbols in the equation make sense in the context.

So, in general, be aware its pro and cons.

Scientists started the field of AI research in the 1950's with the now largely failed quest to produce machines that think. However, they did open the door to making improved individual products that can "learn" how to do their limited jobs better, and they also opened the door to the use of AI in games and in recommender systems such as you see here on Amazon.
This book is the second edition of the successful book by Tim Jones on different facets of AI, how they can be used, and how to write programs that implement the necessary algorithms. The book begins with a short but insightful chapter on the history of AI, followed by a series of chapters, each covering a specific AI technique. The last chapter covers the state of AI today. Each chapter begins with a short description of the technique covered, sometimes including parallels to the real world that are behind the algorithmic choices of the technique. Next, the algorithm is described, and a sample implementation is given and discussed. Last, the author presents examples of problems that can be solved by the given technique. This book basically replaces the first edition, as everything in that book is in this one plus the A* pathfinding algorithm, particle swarm optimization, classifier systems, reinforcement learning, and natural language processing. For several of the techniques variations and tuning opportunities are presented, allowing the reader/programmer to easily adapt the technique to a different problem of a similar type. There are also plenty of illustrations and diagrams, making the material easier to absorb. I highly recommend that you purchase this second edition, even if you already have the first edition. It is a worthwhile upgrade. The author assumes that the reader has already been exposed to the basic ideas of artificial intelligence and is proficient at programming in C. I notice that Amazon does not show the table of contents for the 2nd edition, so I do that here.
1. History of AI
2. Pathfind and the A-Star Algorithm **
3. Simulated Annealing
4. Particle Swarm Optimization **
5. Introduction to Adaptive Resonance Theory (ART1)
6. Classifier Systems **
7. Ant Algorithms
8. Introduction to Neural Networks and the BackPropagation Algorithm
9. Introduction to Reinforcement Learning **
10. Introduction to Genetic Algorithms
11. Artificial Life
12. Introduction to Rules-Based Systems
13. Introduction to Fuzzy Logic
14. Natural Language Processing **
15. The Bigram Model
16. Agent-Based Software
17. AI Today
** Denotes a totally new chapter

Artificial intelligence has grown by leaps and bounds in the last 50 years, but during this time it has also seen a lot of valleys and backwashes, especially in the field of robotics. The unrelenting obsession, driven mostly by military needs, for creating autonomous thinking robots has met with considerable disappointment in the last few decades. This has caused some researchers to distance themselves from the words "artifical intelligence" in order to regain the confidence of funding sources. Thus one hears the words "computational intelligence" or "cognitive science" to describe the field. But sometimes words can accurately describe concepts or properties even they were chosen somewhat cavilierly. "Computational intelligence" could thus be viewed as that branch of artificial intelligence which primarily deals with algorithms designed to deal with large amounts of data, finding interesting and nontrivial patterns thereof.
The content of this book could be viewed as a collection of algorithms in computational intelligence, but also includes topics not usually included in this classification, such as intelligent agents. Indeed, the concept of intelligent agents that the author discusses in the last chapter of the book draws on what he has done before it. The techniques and algorithms that he discusses in these chapters, such as neural networks, genetic algorithms, fuzzy logic, decision trees, and natural language processing, supply the decision-making capabilities for the intelligent agents. These intelligent agents can be viewed as a step towards resolving one of the major issues in artificial intelligence, namely of constructing intelligent software or machines that work in more than one domain. Playing good chess does not mean playing good poker, but developments in agent theory show promise in making expertise in both of these games a reality.

Some of the algorithms discussed in the book have their origins in physics (simulated annealing), biology (adaptive resonance theory, ant algorithms, genetic algorithms, and artificial life), and brain modeling (neural networks). The reader will also be introduced to some of the older methods in artificial intelligence, such as rule-based systems (loosely referred to as GOFAI for "good ole-fashioned artificial intelligence by some researchers). Source code in ANSI C is given for the algorithms, even though the resulting programs are command-line driven. In spite of this, and in spite of no use whatsoever made of Prolog or LISP, all of the chapters in this book will serve to introduce the beginning student or practicing scientist to useful algorithms in AI.

After many letdowns in the past five decades, and also many accomplishments, artificial intelligence is now taking off, and has invaded many different fields with a vengeance. Indeed, financial engineering, bioinformatics, network engineering, elementary particle physics, manufacturing, computer games, and many other fields are making heavy use of intelligent algorithms. A lot of this use has been driven also by the rise of the Internet and the dramatic increase in computational power of computer hardware. Even robotics, the field that has been a source of frustration for researchers in AI, has now shown every sign of finally moving ahead. Without a doubt the 21st century will see the presence of thinking machines. These may not take the form they do in popular entertainment, but in whatever context they are used, one will be able to trace their abilities to the painstaking and patient efforts of the many early researchers in AI. The minds of these machines, however exotic they may be, however advanced they may be, and however they are used, will be products of the incredible originality and skill of the human mind.

Re the comments below, I contact the publisher on the Internet and ask if they had a erratum sheet. In less than a day, a copy was sent to me, and they also have those sheets onsite for most of their publications. Additionally, a second edition was released in October which corrected the identifiable errors. A class act in my opinion.

This book covers a good range of topics but fails to deliver the topics in a clear and concise manner. Not to mention there are many typos (including incorrect and incomplete equations). He also tries to give large examples where a small example would be more suitable. He also does not finish some of the examples (such as the A* method example).

All main directions of AI (Neural Networks, Genetic Algorithms, Ant Colony Optimization, Adaptive Resonance Theory, Artificial Life and many more) are covered in concise and clear manner.
What makes this book outstanding is that application examples which follow theoretic material are considered in greater detail; all applications of AI techniques presented in this book are very well selected and are all insightful.
This book is also very well organized. Short introduction presents an insightful position on the issue; the intro is followed by application example where explanation of issues continues on practical matter.
Algorithms for all topics are written very clearly, and the code is transparent.
In summary, this book is excellent introduction to AI which not only clearly identifies important features of various AI techniques and develops ideas clearly, but also effectively supports it with excellent examples.

This book is a great idea. I am enjoying reading it and working through the examples. I get hooked on each topic, and I find myself absorbed in learning the algorithms and playing with the code. The code is readable, and the author explains the code well.

Unfortunately, there are some errors in this book. I originally complained in my review about not having found any errata on the publisher's site. It turns out they are there now. Last I checked, you get to the errata from the main page, not from the book's page. I am corresponding with the author about some errors that haven't been caught yet. He has been very responsive.

Once the errors get caught and fixed, this will be an amazing book. I don't know of any other quite like it.

It's just amazing how such a 473-page book could cover wide range of AI subjects, and at the same time could deliver easy-to-understand explanations without throwing PhD maths at you. It is just superb how a tech book with minimal math could explain everything!

Every chapter covers one kind of AI, one or two real-world applications and a list of other applications. There're more than enough illustrations for people like me who hate text-only book. A picture is worth a thousand words buddy. Here's my list of AI fields I remembered from the book: Genetic Algorithm, Neural Network, Ant, Particle Swarm, Simulated Annealing, ALife, Pathfinding, A-Star, ATR1, Classifier systems, Rule-based Aystems, Agent-based Software, NLP, Bigram, and Fuzzy Logic.

Examples are written in C language. Comments are plentiful. The codes, though short, pack a lot punches; a whole stack-only virtual machine (VM), for example, was written in less than 100 lines to illustrate how Genetic Algorithm works. Although I'm a VB.NET programmer, I could understand them quite easily.

I recommend this book to any programmer who like to get a big picture of artificial intelligence, who doesn't know where to begin or which algorithm is the right choice. This book should be the first-to-read but not the last, because it touches not far from the surface. I'd say this is the best precursor.

The code is fairly easy to read, and (to date) I have not found any errors in example code... but in terms of the author's ability to write properly, that's a whole different story. Formulas are missing halfs, inconsistencies in variable usage, it's seriously bad. Maybe I got a bad edition (didn't buy from Amazon), but it looks exactly like the one listed (softcover, 2nd ed)..

If you're fairly proficient at C/C++ code, however, you should be able to follow the book ok... just expect to be reading more of the source code than the actual writing, half the time.

It was a required textbook for a class.