News for hackers. Stuff that matters.

smalieslami

Re:Resurgence of Neural Networks on Thursday February 28, 2013 @11:16AM (#5299311)
In fact we're only scratching the surface when it comes to the generative capabilities of deep models. See e.g. our recent work on using Deep Boltzmann Machines to learn how to draw object silhouettes: http://arkitus.com/ShapeBM/
Re:How do we tell truths that might hurt? (E. Dijkstra, 1975) on Saturday December 15, 2012 @09:43PM (#4927126)
I was pleased and surprised to find concise descriptions of several approaches to 'persuasion' on Wikipedia:

http://en.wikipedia.org/wiki/Modes_of_persuasion

http://en.wikipedia.org/wiki/Rogerian_argument

I hadn't thought of 'for grownups' in that way, but I see what you mean. I've now updated the copy.

The resending idea is fantastic. You could, I guess, copy the message into the send message form and click send, but I'll look into adding that as a native function to the UI as well.

The messages are stored on the server. I am a bit nervous about spam, but I plan to deal with that when it arrives. At the moment only one or two of the emails have been 'spam' (messages with only a single word in them).

I'll also look into making some sort of logo ASAP. Thanks for your feedback :)

Re:Software design patterns for Machine Learning R&D on Wednesday August 15, 2012 @03:03AM (#4384958)
This is very good advice.

It almost always makes sense to include Tom Minka's Lightspeed toolbox (http://research.microsoft.com/en-us/um/people/minka/software...) right from the beginning.

Also perhaps Netlab (http://www1.aston.ac.uk/eas/research/groups/ncrg/resources/n...) although it is beginning to get rather dated.