too complex for humans to create the models unassisted? | Hacker News

Hacker News new | past | comments | ask | show | jobs | submit

login

sgt101 on July 17, 2017 | parent | context | favorite | on: Machine Learning Crash Course: The Bias-Variance D...

too complex for humans to create the models unassisted?

ovi256 on July 21, 2017 [–]

Definitely.

Look at the ImageNet ICVLR competition. Hand-built models can't approach SoTA results built with CNNs.

Consider applying for YC's W25 batch! Applications are open till Nov 12.
Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact