Notices tagged with hons

mjog

transform constraints into denials, if a hypothesis entails the body then it is not valid #hons

Monday, 03-Aug-09 05:53:13 UTC from omb →

mjog

why am I still looking for papers? its august! #hons #doomed

Sunday, 02-Aug-09 02:25:03 UTC from omb →

mjog

hunting for a meeting with my supervisor, setting traps at the cafe #hons

Tuesday, 28-Jul-09 00:48:05 UTC from omb →

mjog

coffee, then checking out this new camera. wonder if I can get the pan/tilt working as function of the cognitive units attention #hons

Monday, 27-Jul-09 00:12:40 UTC from omb →

mjog

still looking like the best plan is to convert first-order domain constraint formulas into denials. need a mechanical way of doing it #hons

Tuesday, 21-Jul-09 06:08:22 UTC from omb →

mjog

still not sure how to rewrite arbitrary first order logic clauses into a useful form. a → b ∧ c as a horn clause? to the library! #hons

Tuesday, 30-Jun-09 06:33:00 UTC from omb →

mjog

eureka: minimising ¬abnormal predicates actually makes circumscription seem useful. now to implement that in prolog #hons

Tuesday, 30-Jun-09 03:16:29 UTC from omb →

mjog

basic socket server and command line client for the vision unit is done. tomorrow: prolog client for the cognition unit and some bling #hons

Wednesday, 24-Jun-09 14:16:18 UTC from omb →

mjog

having a hard time pulling a smooth flowing narrative together for this seminar presentation #hons

Wednesday, 29-Apr-09 12:34:25 UTC from omb →

mjog

finally got the #hons seminar presentation started. looking quite good, but too long already. now, fix bike puncture, get to release party.

Saturday, 25-Apr-09 04:40:40 UTC from omb →

mjog

trying to remember probably theory... where's that stats textbook? #hons

Saturday, 18-Apr-09 02:32:09 UTC from omb →

mjog

okay, time to calculate the posterior probabilities of these suckers #hons

Friday, 17-Apr-09 13:23:30 UTC from omb →

mjog

awesome: turns out opencv's hough transform implementation is just what I need to get lines from an edge map. now for step 2. #hons

Wednesday, 15-Apr-09 22:09:57 UTC from omb →

mjog

_Abductive Logic Programming_ is the paper I should have been reading a few weeks ago. #hons

Tuesday, 14-Apr-09 00:03:09 UTC from omb →

mjog

ah, awesome, found a ref and (finally) a copy for /the/ abductive programming paper #hons

Monday, 13-Apr-09 05:52:31 UTC from omb →

mjog

ah, awesome, for a ref and (finally) a copy for /the/ abductive programming paper #hons

Monday, 13-Apr-09 05:52:28 UTC from omb →

mjog

the lack of literature about implementing abduction in prolog is pretty disappointing #hons

Friday, 10-Apr-09 07:48:29 UTC from omb →

mjog

edge extraction is hard, lets go prolog hacking! #hons

Wednesday, 08-Apr-09 13:44:17 UTC from omb →

mjog

hough transform isn't sensitive enough and #opencv's countour finder finds contours, not lines. back to the drawing board #hons

Tuesday, 07-Apr-09 06:00:25 UTC from omb →

mjog

using a hough transform is looking like an interesting way of extracting lines from the edge detection results #hons

Monday, 06-Apr-09 13:32:39 UTC from omb →