AI paper games focus
Hi chaps, Just finished my first year at UWE which included a module on CI. Unfortunately I was referred on a paper and have been issued a new piece of work to do through the summer. I am tasked with discussing the application of CI techniques and thought it would be more relevant to my degree to discuss their application in video games. I need to provide two examples (with discussion) of each of the following: 1) Fuzzy Logic 2) Nueral Networks 3) Genetic Algorythms 4) Collective Intelligence If you have any experience of using these technologies in game development I would be really grateful if you could elaborate briefly on the problems they were used to solve, The details of their use and the results. I only have 2500 words so a paragraph would be sufficient. Thanks guys
Perfection is a product of progress not an alternative.
Well, this is homework so how bout you provide a list of situations in which you could use those techniques and we'll critique it. The purpose of the assignment is for you to think of applications, not for you to get answers from other people on the internet.
I'll give you one comment: none of those techniques are commonly used in game AI. Most game AI is some variant of a simple state machine with perhaps some heuristic oversight. With the exception of 4, those techniques are designed for learning: something that isn't generally feasible within the timeframe of a game play-session. Learning also has the risk of completely backfiring on you. If the player learns that the AI is learning he can force it down a deterministic pathway: "if i do X three times, he'll do Y"
So be creative. I can think of some game scenarios in which it would be useful to employ those techniques. Try the approach of, "what is this technique good for in general" then "what sort of game would present that kind of problem to the AI system?".
-me
I'll give you one comment: none of those techniques are commonly used in game AI. Most game AI is some variant of a simple state machine with perhaps some heuristic oversight. With the exception of 4, those techniques are designed for learning: something that isn't generally feasible within the timeframe of a game play-session. Learning also has the risk of completely backfiring on you. If the player learns that the AI is learning he can force it down a deterministic pathway: "if i do X three times, he'll do Y"
So be creative. I can think of some game scenarios in which it would be useful to employ those techniques. Try the approach of, "what is this technique good for in general" then "what sort of game would present that kind of problem to the AI system?".
-me
Look in your school's library or a local bookstore for the "AI Game Programming Wisdom" book series. In these books, applications of the topics you mentioned to game AI are discussed in detail.
Try these pages for a good start:
http://www.gameai.com/
http://www.ai-junkie.com/
Oh and get the book mentioned above. Read it a while back but it definitely covers the topics you mentioned and how they are used in some commercial games.
http://www.gameai.com/
http://www.ai-junkie.com/
Oh and get the book mentioned above. Read it a while back but it definitely covers the topics you mentioned and how they are used in some commercial games.
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement
Recommended Tutorials
Advertisement