math people?? (matt?)

[HappyScrappy]

Anyone here want to talk about Markov Chains (or matrices)?

I'm using them for my "fake user" of the board that I'm writing.

 

[FrankRizzo]

How about the Gauss Jordan Method?

 

[HappyScrappy]

I'm not up on this stuff as much as I'd like to be - but as far as I know - the Markov is for tracking the probability whereas the Gauss Jordan is... shit, it was stuff with linear algebra - I don't see how it relates to the probabilistic likelihood of word X coming after word Y - than again, I could be stupid...

 

[Sailor_Girl]

1 2 4 . 8 5 9 . 8 10 36
9 5 7 x 3 7 1 = 27 35 7
6 8 2 . 6 2 6 . 36 16 12

 

[MattTheSkywalker]

Ahh fuck!

Gauss Jordan is for linear algebra. Markov.....hmmmm I'll try to check it out.

 

[HappyScrappy]

I'm not sure if this will format correctly, but this is the way I forsee it to work (although this is in theory, the actual code will be different)



i fuck hairy monkeys.

you'd wind up with a table like this

i fuck hairy monkeys
i 0 1 0 0
fuck 0 0 1 0
hairy 0 0 0 1
monkeys 0 0 0 0

if you then processed the sentence:

i eat monkeys

you'd then have a table that looked like
i fuck eat hairy monkeys
i 0 .5 .5 0 0
eat 0 0 0 1 0
fuck 0 0 0 1 0
hairy 0 0 0 0 1
monkeys 0 0 0 0 0



(I have a feeling this ignores anythign beyond 1 space and hence why it is all fucked up)