Open Salon: YouTube ML

YouTube ML

Deep Learning - Geoffrey Hinton

Published on Jul 29, 2015

Filmed at the University of Cambridge on June 25th, 2015

ICLR 2014 Talk: "Zero-Shot Learning by Convex Combination of Semantic Embeddings" by Mohammad Norouzi, Tomas Mikolov
https://www.youtube.com/results?search_query=tomas+mikolov
CLR14: M Norouzi: Zero-Shot Learning by Convex Combination...
by ICLR

2016 Stanford CS224d: Deep Learning for NLP

YouTube 15 videos 18 hours

2016 Stanford CS231n: Convolutional Neural Networks for Visual Recognition
https://www.youtube.com/playlist?list=PLlJy-eBtNFt6EuMxFYRiNRS07MCWN5UIA
https://www.youtube.com/playlist?list=PLLvH2FwAQhnpj1WEB-jHmPuUeQ8mX-XXG
CS231n: Convolutional Neural Networks for Visual Recognition
http://cs231n.stanford.edu/syllabus.html
https://github.com/cs231n/cs231n.github.io
YouTube 15 videos

Ryan Rosario - Sentiment Classification Using scikit-learn
https://www.youtube.com/watch?list=UUOjD18EJYcsBog4IozkF_7w&v=y3ZTKFZ-1QQ

notes:
used Python library Pandas for huge data handling and scalability for multiple cores
http://machinelearningmastery.com/prepare-data-for-machine-learning-in-python-with-pandas/

Scala equivalent - SADDLE developed by the same
The name Saddle is derived from the letters of Scala Data Library. Adam Klein began Saddle development in 2012 while an employee ofNovus Partners, Inc. Saddle was open-sourced by Novus under the Apache 2.0 License in 2013. Saddle evolved from earlier prototypes developed by Chris Lewis, Cheng Peng, and David Cru, and draws on Adam's prior experience developing the pandas Python library.
https://saddle.github.io/

Java equivalent -

SCaVis data analysis package for data analysis

First, it is 100% Java and links many statistical and numeric Java libraries. Second, it is free. And finally, due to convenient design, you can use popular scripting languages (such as Python and Ruby) on top of Java engine. I also did several benchmarks with C++ for data manipulation, and I found it is close to the compiled C++/C.

http://jwork.org/scavis/

http://stackoverflow.com/questions/13811123/data-manipulation-and-analysis-framework-in-java

Analytics + Visualization for Neuroscience: Spark, Thunder, Lightning
https://www.youtube.com/watch?v=cBQm4LhHn9g&feature=em-subs_digest

VC
most popular, evergreen a16z Podcasts of 2016
http://a16z.com/2016/12/27/top-a16z-podcasts-of-2016/

Below are just some of the most popular, evergreen a16z Podcasts of 2016 … Thank you: to all of our guests, from entrepreneurs to policymakers to book authors and more. But we are especially grateful to all of *you* — our audience — who made our podcasts popular by listening to them; sharing them; and including them in your best-of lists, trend write ups, or write ups about entire episodes. Heck, we’re even grateful to those of you offering to pay for them — or hosting dedicated listening parties! It’s true, we do have “a love affair with podcasting” … and though some of you say it’s very, ahem, “un-VC like“, many of these conversations (hallway or otherwise) really do reflect our culture of sharing opinions and ideas; going deep on company building; and thinking out loud — with you — about the trends shaping the future…