(no subject)

[njordsifusansoo]

Ugh. So I was thinking of doing the first option for the project for my pattern recognition class. It basically involves implementing something called LDA from the paper "Discriminant analysis for recognition of human face images". After reading the paper, I realized that a)the paper was written for an audience head deep into image analysis and b)the math goes over my head. So I am now going to do the 3rd option (the second option also involves LDA).
The 3rd option is a computer problem out of the book involving Hidden Markov Models (HMMs). I just spent the last 2 hours or so deciphering what the author of the book wrote about HMMs. Why the hell did he use both a's and b's with alpha's and beta's, and why did he refer to the two sets of data in the problem as w_1 and w_2 when in the HMM section the hidden states are refered to by w_i (with the i being a number).
Well I have wrangeled out the equations I need to know to code the algrotihm they give )as well as gotten and idea what I need to use for inital values). I will implement it tomorrow and hope it all works.

Comments

[saturnvii7.livejournal.com]

Hmm, that first option sounds like a continuation of my not-so-fun linear algebra project from last year :-/