-4
$\begingroup$

I'm a graduate student in Computer Science. For an exam, I have to realize a package in Wolfram Mathematica, with a free topic.

My interests are particularly related to theoretical Computer Science, time and space complexity, and to programming languages.

Any suggestions for a possible project (in this or in other fields...)?

$\endgroup$
  • 2
    $\begingroup$ Could you please exemplify "realizing a package in Wolfram Mathematica, with a free topic"? (I know Mathematica and its packages, but I can't tell what you are seeking.) $\endgroup$ – M.S. Dousti Dec 11 '10 at 16:57
  • 8
    $\begingroup$ @Suresh: I think a question along the lines of "Here's an idea for a project I have, can an expert see any potential pitfalls" would be a stronger question. If it's Mycol's project, why should random internet strangers be doing all the conceptual work? $\endgroup$ – Aaron Sterling Dec 11 '10 at 17:30
  • 1
    $\begingroup$ Also - is the exam at the end of next semester, or at the end of next week? $\endgroup$ – Aaron Sterling Dec 11 '10 at 17:31
  • 2
    $\begingroup$ @mycol: this question runs the risk of being closed if the policy described in the above meta link comes into force (and it's likely it will). You might want to consider focusing the question along the lines of Aaron's comment $\endgroup$ – Suresh Venkat Dec 11 '10 at 20:25
  • 2
    $\begingroup$ Mycol: we wouldn't mind helping, but you do need to be more specific: show us that you've put in some effort by narrowing down a topic or two, and then maybe folks here can comment on what might be useful contributions (again, I refer you to the meta discussion here: meta.cstheory.stackexchange.com/questions/802/… $\endgroup$ – Suresh Venkat Dec 13 '10 at 7:09
4
$\begingroup$

Two possibilities:

  • Encode automata and implement some algorithms for them in Mathematica

  • Implement an abstract interpretation framework for some simple language in Mathematica

Both of these topics have been done many times over, but it would be interesting to see how well/poorly Mathematica deals with the underlying structures.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Regular expression algebra would be a great addition. $\endgroup$ – Thomas Ahle Jan 14 '12 at 0:27
1
$\begingroup$

Implement automatic differentiation method from "Exact inference and learning for cumulative distribution functions on loopy graphs" paper. The idea is to use the tree decomposition of the expression graph to do symbolic differentiation exponentially faster than direct method. Their experiments compare tree-decomposition method with Mathematica's built-in differentiator, and they seem happy to provide code they used for experiments

| cite | improve this answer | |
$\endgroup$
0
$\begingroup$

I would recommend diving into NKS, at least read parts of Chapters 11 and 12 in the book, it is quite likely to inspire you to chart out a reasonable implementation project. Of course, when in need, ask for help here.

| cite | improve this answer | |
$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.