Introduction to Artificial Intelligence
Artificial Intelligence is by now a relatively old field, having originated in the early days of the digital computer revolution. However, it has had a very rocky and turbulent history, going through several cycles of overblown expectations followed by almost equally dramatic swings towards disillusionment and skepticism. In recent years, though, it has matured into a very solid and practical discipline that exercises an ever growing importance across a wide breadth of technologies and professions. We increasingly take speech recognition, handwriting recognition, and natural language search for granted. Basic familiarity with what Artificial Intelligence is, and what tools and techniques fall under its domain, are becoming ever important aspect of a variety of professions and occupations.
There is no shortage of books and resources on Artificial Intelligence. However, most of them fall squarely into two main camps: discursive overviews for the general audience, and highly advanced textbooks requiring deep familiarity with many advanced technical concepts. Ertel’s “Introduction to Artificial Intelligence,” even though it’s pretty technical in its own right, is still fairly accessible introduction to this field for anyone with solid grasp of basic college-level math and computer science concepts.
The book is organized somewhat chronologically along the lines of topics that have historically formed the main organizing principles for the study of Artificial Intelligence — first and second order logic, propositional calculus, PROLOG, machine learning, neural networks. Some of the earlier chapters’ material is a bit dated, and in some cases unfamiliar to students and practitioners in North America. For instance, it seems that PROLOG never quite got a hold on this side of Atlantic. There are a few more or less amusing examples of how quickly technology ages, such as references to Google Video links, which haven’t been around for a few years now. I would have also liked a substantially more material on machine learning and neural nets, maybe at the expense of the earlier chapters. These topics have a lot of practical applications today, and seem to be the guiding paradigms for Artificial Intelligence as a whole for a foreseeable future. Nonetheless, the book overall is very readable and relevant.
One of the most valuable aspects of this book are the worked out examples and numerous (solved) exercises. Working through problems is, by far, the best way to learn any new material, and this book provides the reader with numerous and wide-ranging opportunity to do exactly that.
Overall, this is a very well written and pedagogical book that fills an important niche in the Artificial Intelligence educational literature. Highly recommended.
**** Electronic version of the book provided by the publisher for review purposes. ****
Originally published at www.tunguzreview.com on July 2, 2015.