Artificial Intelligence: A Modern Approach (Prentice Hall Series in Artificial Intelligence)

This is the best introductory review to Artificial Intelligence on the market. It's very well written and organized. There are other books that are better for focusing on one particular aspect of AI, but as a general book this is the best I've seen. If you are looking for a really good introductory textbook to AI that does not completely dumb things down, buy this book. Most of the negative comments about this book come from people stating that it's not a big enough update from the prior edition. While it's true that the entire field of AI has not been completely updated since the last edition, it's also not the case that this book comes out with new editions with the frequency of some Calculus or Economics textbooks where new editions seem to come out purely to ensure students can't buy a used book for the course. The updates are substantial. Whether the new edition gives you enough extra to want to buy it if you own the old edition is a decision only you can make for yourself after spending some time at the website for the book aima.cs.uberkeley.edu. The Kindle conversion of this book is absolutely horrendous. I prefer to buy electronic copies of books if possible so I don't have to carry a heavy hard copy around since I often read while commuting. I would not recommend that for this book, even though at a 1000+ textbook sized pages, it is a pretty substantial book. Fortunately a friend of mine had bought a Kindle copy of the book and I was able to see how bad it was and I bought the hard copy. I recently got an email telling me Amazon was sending out an updated version of the Kindle version of Steve Job's biography because the conversion hadn't been done properly. They really need to do that for this book. Once done it may be a good idea to state on the website that the Kindle conversion has been fixed.

This is by far one of the best books i have ever read. For many years i have struggled with abtuse books on algorithms, programming languages and mathematics. This book is unlike anything i have read before. For the last 20 years, I as a lay person, who has not taken any university classes in AI and programming have had problems moving passed certain levels of complexity in my understanding. Simple things like decision tress and recursion had eluded me in the many books i read, seeming intangible in the areas where they would interface to the real world. Reading the explanations in this book i have finally got it. These authors through scenarios, explanations and code (replicated in many languages on the companion website) explain decision based structures and there use in agent design ,assuming very little along the way, outside of an elementary understanding of programming and logic, thus not losing me the reader. They gave me the building blocks and showed me how and why i would want to assemble them. The code in this book is both beautiful, elegant and succinct all at the same time. The hours of exploration and learning i got from these samples alone were worth the price of the book. If you are learning to program -- Read this book. If you interested in decision and value based reasoning -- Read this book. This is not a short book, but it is well worth the time investment. I plan on having it done within the year. I have learned so much about what i realize i knew so little of before, but I am so much more capable now than i ever was before, and that is because of this book.

This review is organized into two parts. The first part gives a general overview of the textbook and my thoughts about the content, while the second part gives a summary of the problems I've encountered while reading the Kindle version. This book was made by multiple authors. Fortunately, the book's definitions and summaries are generally coherent. It has a through coverage on the history of AI, the definitions (those that are controversial have been noted as such), and provides psuedocode of many algorithms. Similar to other textbooks on this topic, it does not provide run-throughs of those algorithms. The organizations of the book can be improved. I have some problems with the way they organized the material on the second part of the book (II: Problem Solving). Chapter 3 is named "Solving Problems by Searching", but I didn't realize all they did was on classical searches in Chapter 3 until I read Chapter 4. They could have explained why uniform-cost search is uninformed search even though it has a cost function better than they did in the textbook. A better naming scheme for those chapters would be "Ch3: Global Search ", "Ch4: Local Search/... (I would suggest breaking Chapter 4 and put it into different places.)". I bought the Kindle version, and it had many formatting problems. Most of them are minor (such as missing a "space" between two words, images appearing in different locations, image sizes too small, etc.) However, I had caught one specific error in the second part of the book (somewhere between chapter 3-5) where the fact is exactly the opposite than the one written (forgot to add the negation?) I should have marked and noted it so I can write down which error it was in the review but I was too lazy... But it's a good purchase overall!

This is indeed one of the best books on Artificial Intelligence (AI). It is comprehensive in that in its 1,145 pages it covers and effectively explains concepts and practices; It is authorative as it is written and edited by some people who are pioneers in AI especially Peter Norvig; Being the 3rd edition it clearly updates and augments another 20% to the 80% of the 2nd edition with updates. With the massive number of reviewers and contributors, the book presents a balanced, realistic, and and down-to-earth understanding. The biggest attraction of this book, unlike many other books on this subject is that: 1. It avoids hype but humbling explaining things, and 2. It does not force you to understand things by referring you to some code fragment that is difficult to understand or execute. I liked the fact that this book has a nice and logical organization. Contents: I. Artificial Intelligence, II. Problem Solving, III Konwledge, Reasoning, and Planning, IV. Uncertain Knowlege and Reasoning, V. Learning, VI. Communicating, Presenting, and Acting, VII. Conclusion. Excellent work!

When this book first arrived, I had a lot of buyer's remorse. I wanted something to read on the side, not as a textbook for a class I'm taking or anything like that. I wasn't expecting it to be this big. How could I possibly read all this? However, it's big for a reason. This book is filled with every possible thing you might want to know in the field of AI. It is all-encompassing from what I've seen. It also explains things very well in my opinion. Things are clear. You can clearly go through it, and although the math might need some note-taking, the book generally makes it easy for you to understand with examples. If you want to start delving into AI, machine learning, etc, but feel daunted about all the different crazy terms they use (adverserial networks, loss functions, overfitting, etc.), this is a good book to introduce to all this. It certainly did for me (Chapter 18: Machine Learning)

The book is awesome. I'm using this for a class I'm taking right now and would recommend anyone interested in AI read this. Additionally, the book is a great reference. Topics are highlighted in the margins; which simplifies finding sought information. My only complaint is with Amazon. I ordered this book "NEW" but received a very obviously used item. The book has slightly dented corners, split seams in the inside cover, a cracked spine, and obvious curves in pages where the book lay open for extended periods, and was torn in the upper corner of the front cover. Given the condition, I would have returned the book immediately. However, as said, I am taking a class and have assignments due from this book and so can not send it back and can not be without it. I wasn't sure how to contact Amazon to correct this issue as this is my first time purchasing from them. While I would recommend this book, I likely will not purchase from Amazon in the future.

As a textbook, it's fine. It covers a wide range of concepts and topics in the field of artificial intelligence and is a good resource to have around. My rating is mainly based on the state of the Kindle version. There are some issues I have with it that keep me from enjoying it further. The biggest problem is the overall format. Pages aren't just digital versions of the physical pages in the real book. They're broken up into thousands of screenshot-like segments. You can't scroll "up to the top of a page" or "down to the bottom" because each page segment is only as big as your screen. This would be fine if it were a much smaller book. My tablet counts 33,261 page segments when I use it, so you can see how this might be cumbersome for such a large book. The effect of the above problem could be reduced a bit if the table of contents were complete. As it is, the table of contents is very limited in what section it allows you to "jump" to. The book is broken up into sections, each of which is covered in a number of chapters, as one might expect. But unfortunately, the Kindle version only lets you jump to the first chapter of each section, not any chapter in the book. This means that you either have to swipe through hundreds of page segments to get to the spot you want, or you can use the slider at the bottom of the screen and hope you land near where you want to be (the slider won't give you any additional info about where you are, either). I suppose you could swipe through it ahead of time and place book marks at each chapter header, but that shouldn't be something a reader has to do to navigate a book by its own chapters. The only other things I can think to mention have to do with resizing the screen. It's not very smooth at all. In fact, there is usually a noticeable delay when you try to resize a page segment, probably because it has to re-calculate all the page segments to fit the new zoom scale. I'm sure this could be improved if each page were just a digital version of its physical counterpart. The other problem is that the images in the book do not resize with the text. This is an especially troublesome problem for this particular book, since there are a number of intricate diagrams and images of pseudo code algorithms that are very difficult to read at this size. I can't even read some of them, and my tablet screen is larger than 10 inches. This is actually the second Kindle book I've purchased. The first one (Introduction to the Theory of Computation, by Michael Sipser) was near flawless. Each page was a digital representation of its physical counterpart, the table of contents was complete, and resizing was very smooth. I bought this AI book in Kindle form hoping for a similar experience, but I guess not all the kinks have been worked out yet. It honestly comes off feeling like a beta version that isn't quite finished. But given my experience with that other Kindle book, I know improvements can be made to this one. All that being said, I can still use the Kindle version of this book. It's just very inconvenient. If you're looking to acquire this book, I'd recommend going with the physical version, for the time being. Yeah, it's another large textbook to haul around. But it will be a lot easier to navigate and use.

Great book ! 1130 pages and yet to the point. I haven't read the first two editions so i cannot judge if there is a lot of new stuff in this 3th edition, but i feel like i am able to make an expert system myself, having read the book. Not for the layman though. You need to have an engineer's or mathematical mind to grasp everything. The book really covers everything from the neat's point of view. In the heart, i am a scruffy but what this book has taught me, can certainly do no harm as it is a real overview of all the scientific techniques involved in making expert systems and other applications in AI.