Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
latenightcoding
on March 5, 2018
|
parent
|
context
|
favorite
| on:
A week-long programming retreat
New regularization techniques to stop these massive NNs from overfitting is also one of the main factors for the current deep learning explosion.
mafribe
on March 5, 2018
[–]
Which regularisation techniques have you got in mind?
cdancette
on March 5, 2018
|
parent
[–]
i'd say dropout, batch norm
Consider applying for YC's Spring batch! Applications are open till Feb 11.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
