Log in

No account? Create an account

Dreaming in graphs and pictures

« previous entry | next entry »
Apr. 12th, 2007 | 06:34 pm
location: the lab.
mood: busybusy

To write something that is not related to relationships...

I accidentally left my jump drive in a jeans pocket recently and sent it through the wash. Apparently, this did not hurt it, as it's still being plenty effective for transferring data (in the form of .zer files) from one lab computer to my laptop. Yay for solid-state components.

In other news, bees are fascinating creatures. They have very neat visual systems. (Relatedly, I've been surprised by how often oxyopia has looked interesting lately and how excited I've been about going. I remember being a first-year and feeling overwhelmed by half the talks and bored by the other half. Now I genuinely find them fun and entertaining and interesting. Ah, affirmation of self).

Also (based on today's talk, which showed a way that a neural network with a specific set of Bayesian priors could "see" illusions almost identically to the way humans seem to) finding myself suddenly motivated to learn about neural networks. (Thesis first... thesis first... thesis thesis thesis...)

At least I've outlined the introduction. So I think I can do a good bit of fleshing that out tonight. Same thing tomorrow for the methods. And by the time I've done those two, I am likely to have actually a lot written. Then I can throw myself headlong into analysis and see what my data are actually doing and what it all means! (Read: the fun part of science).

ETA: After ... probably 4 hours of solid work, I have a little over a page SS typed that I'm happy with. Of course, I'm (at this point) maybe between 1/4 and 1/3 of the way through my outline, so that's probably about right.

I don't want to think about how many pages of writing I need to do in the next week, between intro/methods draft and journal article draft.

Though the latter worries me less as it will be met with a crew of supportive people and at least half of the writing involves cutting apart and reframing the stuff that I've already written, which is, um, a lot easier than trying to write new stuff from scratch.

Link | Leave a comment | | Flag

Comments {4}


(no subject)

from: stellae
date: Apr. 13th, 2007 07:54 am (UTC)

Oh yes.

Today I am proud of the fact that I have a working definition of "Bayesian priors." It might be the only thing I learned from Stats from Hell (otherwise known as the course that resulted in negative knowledge).

Of course, it helps that I'm no longer required to go... so I can be fairly secure in being just not very interested in corneal disease or diabetic retinopathy and skipping those weeks...

Reply | Parent | Thread