Has anyone here read this book and if so, what do you think of it? I bought it because various people told me it was one of the best books available for AI however upon reading the first few chapters it seems really heavy going, especially mathematically. Is the field really this difficult even at the basic level or are there other books that do a better job explaining things in laymans terms? What level of math is required to understand a lot of the AI algorithms etc.? My knowledge of math only extends to A-Level level which includes Calculus, Matrices, 3D Trigonometry etc. Is that enough or will I need to teach myself more? Perhaps I might be better buying a PrimaTech book and then going on to AI:AMA when I have a decent understanding of the subject?

Yes, I am reading that book right now, and I think it is one of the best books on the subject, and if you have that level of math knowledge and you still don''t get it mathmatically, then give it up. Try Discrete math, that''s all AI is. And AIAMA is NOT very rigourous in math. Read it again.

My Kung Fu is stronger.
neo88

I understand a bit of the math in it but it''s the sigma(the funny E shaped symbol) stuff that''s really causing me trouble, the only time i''ve come across that symbol is in standard deviation.

Do you know any words I can type in to Google to learn about sigma with the numbers and letters above and below the symbol? It''s hard to find stuff when I''m not sure what I''m looking for.

I guess I''ll just need to keep re-reading it until I learn it.

Sigma just means "sum." It means "add up all of these values." It''s like an integral sign, but for a finite set of values. The first hit I got for a google search for "Sigma notation" turned up this which looks like a rather simple introduction.

I consider AIMA to be a good reference for those who know enough about the subject. Why not get yourself a practical book? It''s often easier to understand when the theory is applied.

If you have a good programming background, then why not try my book: AI Game Development. It''s quite accessible to game developers.

Alex

---

AiGameDev.com

We had sigma stuff in highschool sometime, think it''s pretty basic stuff mostly. Did a quick google and got this, seems to be a good explanation.

Get a book on Descreet(sp) Structures, or if you''re in college, just take the Descreet Structures class (it should be a 200 level course). It goes over the funny notations, and give a pretty good overview of math theories.

Well, if you need to learn about Maths, I think you could go get a look at MathWorld
Although if you find the concept of Sigma something out of your league, I have a feeling that maybe you are a bit early to start looking at classic AI stuff...
What school level are you at, out of curiosity ?

---

Sancte Isidore ora pro nobis !

[edited by - ahw on March 2, 2004 2:34:59 PM]

I''m at A-Level level, I understand the Sigma stuff now it''s just that we weren''t taught it at school. I wasn''t sure on what to search for so I found it difficult to find any information on it so I could learn.

But looking through the book I now understand a lot of the formulae . Thanks.

AI:AMA is an excellent text book for AI, however it isn''t the be all and end all for AI. In particular, being a text book, it gives only a cursory introduction to the topics covered (albeit a good one)! There are more specialist texts out there on each of the subject areas covered in AI:AMA and after reading through that text, you should seek out these more specialised texts.

As to the level of math required for doing AI, that really depends on the area you get into. Knowing elementary mathematical methods (calculus, trig, analytic geometry, linear algebra, etc) is a must. Knowing basic probability theory is essential for many aspects of AI in current use. I would suggest that you get a hold of a good set of basic maths texts. A bare minimum would be 3 books: Calculus & Analytic Geometry, Linear Algebra, Introduction to Statistics/Probability theory. If you''re doing AI you''ll certainly utilise them a lot!

Cheers,

Timkin