Machine Learning with TensorFlow (GDD Europe '17)

TensorFlow is the most popular open-source machine learning framework in the world. In this video, Andrew Gasparovic gives you an introduction to TensorFlow and goes through new things that have been developed. He also covers some use-cases demonstrating how TensorFlow is used by Google as well as out in the real-world.
Check out the ‘All Sessions’ playlist for the rest of the talks that were given at GDD Europe ’17: goo.gl/6BFfHv
Subscribe to the Google Developers channel: goo.gl/mQyv5L

[@sreelalts1892]

Thank you. Really interested in TensorFlow 😊

[@alexandrugheorghe5610]

Most excellent, thank you for this talk!

[@XuefeiPei]

Nice presentation!

[@mirandaliu9961]

Great presentation.

[@MaxRovensky]

If they changed cats and dogs to hotdogs and not hotdogs this presentation would be so much more interesting

[@richardf9112]

great presenter!!! thoroughly enjoyed this

[@noli-timere-crede-tantum]

One detail for those following the code snippet at 6:25 is that loss function should be (1/2) * tf.reduce_sum(tf.square(lin_model - y)) and x_train and y_train should be np.array() instances. Else that code will not run =)

[@dciug]

Emm, not really. Works perfectly on tf.__version__ == 1.2.0-rc2

[@zhuuxi]

The 1.25x speed option works just nice for me.

[@mamokilo]

1.5 is much better

[@CerebrumNigthmare]

good one

[@dhavalnarendrashah7938]

if i give Cat and Dog in a single image does it return Cat and Dog as outcome?... for e.g. how it will handle a single image with various animals... as input data is not fixed

[@DaoudMousselmal]

Hello can we get the slides of the presentation ?

[@eybustt3243]

with tensorflow can we develop AI chatbot...??in windows platform

[@70ME3E]

omg the cucumber farmer dude! so awesome,,,,

[@MDMAHBUBALAM]

why tensorflow is not working with python 3.7 ?

[@sarthakairan6548]

it does not support 3.7 at the time of writing this comment use 3.6

[@noli-timere-crede-tantum]

Don't mean to be a hater, but that Udacity course on 31:56 is literally just a bunch of hand waiving...

[@lordnaive]

Rodrigo Silveira yeah truly.. I feel its a waste of my money.

[@AsifMehedi]

I see two from Udacity there. Which one are you talking about?

[@AsifMehedi]

Which one you thought was a waste of money?

[@lordnaive]

The ML one

[@AsifMehedi]

দাদা, I think it depends on one's expectations going in. I've just started the Machine Learning Engineer Nanodegree and I'm enjoying it so far. But that may be because I'm doing it more as to refresh and solidify my understanding rather than learn new material. I already knew most of these things from books and other courses.

[@vijayabhaskarj3095]

The Deep Learning course from Udacity(not the nanodegree one) is the worst course on deep learning on the internet!

[@rujiahao4284]

lol

[@Abhishek.Saini28]

How can we use TensorFlow(Machine learning) in AEC (Architecture Engineering Construction) industry?

[@ZEFIRSATELLITE]

I believe You can send your Architecture data to train tensorflow model so its can predict different characteristics of Constructions :-)

[@peristiloperis7789]

First thing is to identify a specific problem you want to solve in that area.

[@SuenJason]

Thanks!

[@deeplearningpartnership]

Nice.

[@fallrepublicwhydonttwenty3833]

Google😀😃 G.O.O.G.L.E

[@kidslearning8996]

1.5 speed

[@Odyvers]

Tenssor yeah what is about?

[@chanlito4896]

I'm screwed I don't know python.

[@rujiahao4284]

yes, u are screwed

[@namelastname4077]

8:55

[@theoneandonly6316]

if you know any mainstream programming language, you can learn python really fast. python is super easy to learn

[@peristiloperis7789]

I'm about to start learning it. We still have time to catch up.

[@FirstNameLastName-fv4eu]

Wasted my time!!!

[@namelastname4077]

not for developers

[@70ME3E]

but ye