search

openworm / behavioral_syntax

behavioral syntax analysis based on the paper of Andre Brown
Other
5 stars 1 forks source link

automated search for multimodal distributions #11

Closed ghost closed 7 years ago

ghost commented 8 years ago

In order to find discrete behaviors/biological signatures in an automated manner it might be interesting to write a program that automatically looks for multimodal distributions across all continuous features(turning angle, pauses...).

ghost commented 8 years ago

I could use PyStan or some other MCMC library for this. Here's an introductory tutorial: https://github.com/twiecki/WhileMyMCMCGentlySamples/blob/master/content/downloads/notebooks/MCMC-sampling-for-dummies.ipynb

MichaelCurrie commented 7 years ago

@AidanRocke is back, baby! Nice to see you again. :)

ghost commented 7 years ago

Yes. I realised that the first time I worked on 'behavioural syntax' I was trying to do too many things at once. Now, I think I have a simpler solution.

Instead of producing lab reports which is definitely possible, I'll build a web-based application that will simply produce three possible answers given image/video input: 'Yes', 'NO' or 'Maybe'.

That's something most humans would understand. :) My hope is that it will inspire new people to contribute to OpenWorm.

I'll try to take small steps on it this week but next week and the coming weekend is when I'll truly be able to make important changes.

Aidan

On Mon, Jan 30, 2017 at 12:19 AM, Michael Currie notifications@github.com wrote:

@AidanRocke https://github.com/AidanRocke is back, baby! Nice to see you again. :)

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/openworm/behavioral_syntax/issues/11#issuecomment-275958359, or mute the thread https://github.com/notifications/unsubscribe-auth/AFD1LiEeVBe19qtL-HgmRKiD9bHjX35Uks5rXSyngaJpZM4G01pQ .

MichaelCurrie commented 7 years ago

Sounds like an interesting approach! Starting with something simple is usually better, I agree.