Prerequisites and False assumptions and claims on DL

[EmbraceLife]

Prerequisites and False assumptions and claims on DL

8:51-10:23
Prerequisites and False assumptions and claims on DL
- you can’t do DL with 1 year python and high school math
- DL is a black box
- interpretable models
- make sense parameters with viz
- Need too much data
- transfer learning don’t
- sharable pretrained model
- Need PhD to do cutting edge research
- fastai & keras libs, MOOCs
- Jeremy only has a philosophy degree
- great application only for vision
- speech, tabular data, time series
- needs lots of GPUs
- one in cloud is enough
- more for large projects
- Not really AI
- not a brain
- but doing real world useful things
目录 Lesson 1 知识点分解 #63