FastAI Study Group

Meetup

https://www.meetup.com/_ChiPy_/events/264312891/

https://docs.google.com/presentation/d/1x4sY1L_BTwl1fyjeRcFODq-7HIYSULFPZyiyIwAyB4c/edit#slide=id.g324fda911d_0_95

https://github.blog/2018-09-18-towards-natural-language-semantic-code-search/

Starting point for Learning FastAI - Course

https://course.fast.ai

FastAI Forum

https://forums.fast.ai/

Lessons

https://github.com/hiromis/notes

Kaggle Lessons

https://www.kaggle.com/sundance14/shankar-fastai-v3-lesson-1/edit

https://www.kaggle.com/init27/fastai-v3-lesson-2/notebook

FastAI v3

https://course.fast.ai/

FastAI - youtube

https://www.youtube.com/watch?v=fvUcUpUmth8

What is FastAI

https://www.developer-tech.com/news/2018/oct/03/fastai-python-library-ai-simpler/

What is Kaggle

https://en.wikipedia.org/wiki/Kaggle

Ready to run: “One-click” Jupyter

These are the easiest to use; they’ve got all the software, data, and lessons preinstalled for you. They’re a little less flexible than “full servers” (below), but are the simplest way to get started.

  • Salamander (instant approval; no installation required; includes full terminal access; $0.38 an hour; $75 free credit for students)

  • Colab; (instant approval, requires minimal installation, free)

  • Paperspace Gradient; (instant approval, no installation required, $0.59 an hour; $10 free credit)

  • SageMaker; (requires wait for approval, not quite “one click”… but pretty close, $1.26 an hour + storage)

  • Kaggle Kernels; (Instant Launch, No setup required, Free, not always up to date and not as well supported by fast.ai)

  • Floydhub; (instant approval, no installation required, $1.20/hour + $9.00/month (100GB storage), 2 hours free credit)

——

Feature Engineering & IBM Watson Studio

https://www.meetup.com/DuPage-Business-Analytics-Meetup/events/265543397/

https://dataplatform.cloud.ibm.com

Dataset

https://github.com/Meaad96s/datapreparation_titanic/blob/master/titanic.csv

https://ieeexplore.ieee.org/document/6771123?anchor=relatedarticles

Previous
Previous

RESEARCH: Jeremy Howard Interview