search

diegocasmo / ml-project

ML Project
0 stars 0 forks source link

Report - Explain data decision tree and include an image of it #24

Closed tristaaan closed 6 years ago

tristaaan commented 6 years ago

One thing that’s been bothering me a bit is that we don’t really know much about the morphological output we’re trying to predict. It would be a good addition to the data section as it would reinforce why a deep cnn is the right choice for this problem.

ankurshukla03 commented 6 years ago

More like the question that why deep cnn is right choice for image classification??

It would be a good addition to the data section as it would reinforce why a deep cnn is the right choice for this problem.

tristaaan commented 6 years ago

Yes image classification, but also how it specifically applies to this problem where we’re looking at morphology of galaxies and what these 37 attributes are and why a cnn is good at identifying those attributes. (oh what a bad sentence)

On Sat, May 19, 2018 at 08:41 Ankur Shukla notifications@github.com wrote:

More like the question that why deep cnn is right choice for image classification??

It would be a good addition to the data section as it would reinforce why a deep cnn is the right choice for this problem.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/diegocasmo/ml-project/issues/24#issuecomment-390384080, or mute the thread https://github.com/notifications/unsubscribe-auth/AFN5m2RAH8AVIRXcjE7xPqN64ULVaneHks5tz76YgaJpZM4UFkqH .

-- -Tristan Wright