theCricketer

Piero is now at Stanford :). Piero is definitely worth listening to. He's been in the weeds, knows the details and yet keeps the bigger picture in mind. Also one of the nicest people around.

The videos from the winter 2017 offering are freely available on YouTube.

https://www.youtube.com/playlist?list=PLqdrfNEc5QnuV9RwUAhoJ...

The assignments are at the class webpage too.

I took this class and can vouch for it. They update the class every year to go over recent research - not an easy task in such a fast moving field. For example, this offering covers the Transformer architecture which has recently been used to obtain state of the art results across a wide range of NLP tasks.

Since we're on the topic of tutorials to understand neural nets and modern deep learning, I will throw in Michael Nielsen's excellently written free online "book" on neural nets. It's really a set of 6 long posts that gets you from 0 to understanding all of the fundamentals with almost no prerequisite math needed.

Using clear and easy to understand language, Michael explains neural nets, the backprop algorithm, challenges in training these models, some commonly used modern building blocks and more:

http://neuralnetworksanddeeplearning.com/

This book opened my eyes to the power of textbooks written in such easy to understand, clear style. Bet it took repeated revisions, incorporating feedback from others and hours of work but such writing is a huge value add to the world.

I'm not too knowledgeable on how these deals work, but figured someone on HN would know:

A quick Google search shows that Twilio's market cap is currently $7.4 billion. Does this $2 billion "all-stock" transaction mean that they are giving away over a quarter of the company to pay for this acquisition? Or how else should I read this?

> Auto-encoders are overplayed, mostly because they're a pretty easy intro ML project.

I think you mean "normal" autoencoders, like denoising autoencoders or the identity autoencoder that are used for feature learning. Note that variational autoencoders are not really autoencoders in that sense. They are called “autoencoders” only because the final training objective that derives from the probabilistic setup does have an encoder and a decoder, and resembles a traditional autoencoder.

Traditional autoencoders are the common intro projects used for representation learning and to bootstrap other networks, not variational autoencoders.

The insight that made it possible for me to grasp VAEs was digging into the probabilistic setup that leads to this formulation. The neural networks are "just" a powerful function approximators applied on top of this probabilistic framework.

"Without a warning, cause or right of reply the board fired me".. There is more from Tim Kentley's twitter: https://twitter.com/TimKentleyKlay

He just posted a bunch of messages he got from his team at Zoox. Seems like a lot of employees really liked him and his leadership.