Fixed power range to start at 1

pkmital / CADL

ARCHIVED: Contains historical course materials/Homework materials for the FREE MOOC course on "Creative Applications of Deep Learning w/ Tensorflow" #CADL

https://www.kadenze.com/courses/creative-applications-of-deep-learning-with-tensorflow/info

Apache License 2.0

1.48k stars 732 forks source link

Fixed power range to start at 1 #74

Closed alvaroingold closed 7 years ago

alvaroingold commented 7 years ago

Fixed the power range in section Over vs Underfitting, in order to start at value 1. This avoids having a 0th expansion which introduces additional bias. This issue was raised in the session 02 forum here

alvaroingold commented 7 years ago

Closed the fix and merged with another one also regarding section-02.ipynb