AI Project Examples

In theory, algorithms are evaluated by performance measures such as time and space complexity. In AI these concerns are secondary to how the algorithm performs relative to people. You should consider how you will evaluate your program from the outset. You should also pay a lot of attention to the user-interface. Problems or confusing aspects of your interface are most easily found by those not involved with the project. The suggestions below only provide some examples. You might look through any AI text for more ideas. It's best to choose a topic that you are very interested in. All code will be done in Java. Each object should be documented with who wrote the code. Moreover and very important, each method should be documented with time and space complexity when it is not constant. Usually this is not difficult, but occasionally it is very hard. If you can't do it, say so or discuss it with me or the class. In the first lectures I describe some possible projects. You may do your own project, but it needs to be well-defined, demonstratably doable within one quarter, involves some AI method, and approved by me.