Hi,

@Toby, I considered trying to write real genetic coding but honestly I only had this weekend. =)  So I had to settle for randomized selections. However, its possible to derive quite a few simple and perhaps moderately complex equations using my approach. However I need to make a few changes to the approach….and I’m hoping to have something by the end of the day. It would be nice to include basic trig functions, square roots, natural logs, etc.  My current approach doesn’t allow that…but I’ve got a few ideas.

Do you write programs?

@ CR, a systematic approach would reduce computation time but would require a lot of programming related to knowledge of the operators. One example…the number of ‘(’ should equal the number of ‘)’, equation shouldn’t end in a ‘(’, parenthesis should be paired, etc.

The randomized approach does have the advantage of allowing the addition of new operators (familiar to Excel) without coding any heuristics.

However, I’m not ready to bail on this yet. I’ve come up with some ideas today in church (I know, should have been listening) to allow for more complex expressions.  I’ll give it a go this afternoon.

Regards,
Chuck

It would be better for the program to figure out for itself patterns in the structure of equations.  And that is one of the big problems of AI—determining when a new pattern is being presented and acting accordingly.

Hi Toby,

It would be better for the program to figure out for itself patterns in the structure of equations.

I’m thinking of some scrabble games I’ve played against my family. There have been times where I didn’t have a clue. However, I can state two observations:

1. I randomly shuffled tiles initially….crossing my fingers….and hoping something would present itself.

2. In the shuffling, I did adhere to a few patterns that could be expected (ed, gh, ie, ai, sh, ch, etc.)

So, it is possible (with coding) for the program to increase its knowledge base of ‘common’ math expressions.  For example.

a. Calc area of a rectangle:  A = H x W   (or abstract Result = A x B)
b. Calc voltage:  V = I x R   (or abstract Result = A x b)

So in future problems, it could apply the abstract concept Result = A x B.

I think I may be able to do this….or not. =)

Regards,
Chuck

So in future problems, it could apply the abstract concept Result = A x B.

It’s been done by others so there’s no reason why you can’t.

It’s a “symbolization” or maybe a “metaphorization” process.

Some concepts that could be given to the bot when it is born could be “symmetry”, “pair”, “adjacent”, etc.  Check out the CopyCat project by Hofstadter and Mitchell.