Options for ML part of final exam

There are three possible projects for the "final exam" requirement of CSCI 394.

• Verifying arguments automatically. This explores "automatic theorem proving", specifically having you write a program that performs Game 2 problems from the DMFP course manual, chapter 2. Book section. Code for the project. You do not have to do part 2.D.
• Unification and resolution. This explores the field of logic programming, building a system that makes inferences from a set of relations. This is a classic "AI" topic, so if you're sore that we're not offering the old AI course any more, this gives a taste of some of the algorithms behind it. I suspect this is the most difficult of the three options. Book section. Code for the project. You do not have to do part 4.D (but do 4.E).
• Modelling (mathematical) functions; doing calculus. This looks at representing functions (polynomials, rational functions, step functions, exponential functions) and implementing operations on them (evaluating, differentiating, integrating). If you have taken Programming I and II, this will be analogous to things you have seen or done in Java and C (those of you who have done a lot of object-oriented programming will appreciate parts of this project for how they bring in OO concepts to ML, but you do not need any Java or C background to do this one). This one is a little longer (number of parts to the project) than the others, but I think it is easier than the other two. Book section. Code for the project.

For any of these, you can get extra credit for finding mistakes in the book excerpts or suggesting improvements to the presentation.