Deep Learning With Python & Tensorflow - PyConSG 2016

Speaker: Ian Lewis

Description
Python has lots of scientific, data analysis, and machine learning libraries. But there are many problems when starting out on a machine learning project. Which library do you use? How can you use a model that has been trained in your production app? In this talk I will discuss how you can use TensorFlow to create Deep Learning applications and how to deploy them into production.

Abstract
Python has lots of scientific, data analysis, and machine learning libraries. But there are many problems when starting out on a machine learning project. Which library do you use? How do they compare to each other? How can you use a model that has been trained in your production application?

TensorFlow is a new Open-Source framework created at Google for building Deep Learning applications. Tensorflow allows you to construct easy to understand data flow graphs in Python which form a mathematical and logical pipeline. Creating data flow graphs allow easier visualization of complicated algorithms as well as running the training operations over multiple hardware GPUs in parallel.

In this talk I will discuss how you can use TensorFlow to create Deep Learning applications. I will discuss how it compares to other Python machine learning libraries like Theano or Chainer. Finally, I will discuss how trained TensorFlow models could be deployed into a production system using TensorFlow Serve.

Produced by Engineers.SG

Help us caption & translate this video!

[marcinbrdys]

Cool video ... difficult area that even professors find tricky to explain. Well Done Ian Lewis you hit the 10/10 mark can we Google that :)

[michaelcharlesthearchangel]

Thank you for the tips on setting the weights for a train step or rather defining the ssirCross parameters and quantum functions of a diagrammed (meta programmed) "Train Car" or "Train Engine" or "Train Fuel-car" or "Train Clause::Caboose that runs on the Choice Train tRacks represented by multi+axiom hypergramming.

[Fransamsterdam]

One thing I know for sure. If you play Google's quickdraw you are learning how to draw, rather than the program will learn how to recognize your drawings.
And regarding self driving cars: The government in the Netherlands is now already making rules how to build new roads so they will be suited for self driving cars.
So in a few years we have artificial intelligent roads suited for a simple algorithm.
It's a bubble.

[cherryentertainment]

please help to me . i want deeplearning tools (deeplearning 4j). how can i download it.

[heldervelez]

he is a developer talking to developers

[glenneric]

Thanks for the data!  I hope they can one day apply AI to removing excessive 'uh's from videos. (ducks his head).

[GuruKal]

I don't think this guy knows enough about the field to be representing Google at a conference

[HellTriX]

This video would be half as long if he stopped with the "uh, umm, uh,
Humans need far more back propagation training.