วีดีโอ

Very useful talk. John is great

Has a functioning tool come out of it that's available somewhere? that can extract road edges/curbs?

the cure for cancer is: what source of energy can all cells survive on except cancer? Answer; Ketones. Solution: starve the cancer by depleting glucose via a low carb high fat diet. Many people have now cured their own cancer through diet alone.

Very well spek

This is mathematically so simple but yet so profound. Absolutely great!

Is there any link to the example that was demonstrated?

great discussion

Can't the GMM be solved using SVM? What's the difference : separability vs higher order relationships?

Wow finally someone who understands this problem

David Brin predicted the AI empathy crisis with Google's LAMBDA.

Awesome talk - thanks

what happened to fakerfact ?

This is really well presented and provides high quality useful information on applications of AI.

Thanks for the video.

very interesting and informative presentation thanks

Annoying to hear this talk. Come on man. This is not race. There is no point in packing everything in 20 mins if the audience not going to understand it. Quality is important than quantity.

@9:20 Speaker says "I would probably not like the colleague's earthquake post" and I see Mark Zuckerburg liked that post. lol

Leslie Smith is a great speaker in addition to having contributed significantly to deep learning

where do you get the dataset please share the link

a 2% performance boost after all that?

Very good

Great talk!

FAIL!

This has some great insights and learnings. Thanks 🙏🏻☺️

Amazing work and interesting use of embeddings, with a strong privacy as profiles are not part of the training data. Thanks for sharing ! I'd be curious to know what is the current state of Tinder recommendation algorithm.

Thanks for such an insightful talk, par excellence

nice

awesome. he also gave a great talk on reinforcement learning, which I found on TH-cam.

my self prefer tensorflow

Thanks for this awesome pres !

i just cash out 15,000 USD paypal hack transfer from VASTLINKSHACK>COM guys their service are extremely awesome!!

Egonet

Although it was done in 2016, but still found useful nowadays

Interesting Idea of fairness optimization not much proof it works, also no explanation how optimally groups are formed that too involves fairness questions i.e formation of groups.

The point of feature engineering for machine learning, like classification, is a quite limited issue compared to the things the brain does. So it does not really make sense to oppose generic learning models to single purpose ML tasks. That's the whole point why we look at brains. Since nobody rejects that data-driven and prior must be combined, the question is *which* priors (or innate knowledge) and not *if*.

I'd love to know what Evan said while the audio cut out.

Very interesting and necessary work! Thanks for doing that

what happens when the funciton that we sample ist of stochastic instead of deterministic nature? would gaussian processes( or another surrogate model) and thus bayesian optimization have a way to deal with that? EDIT: im referring to what he scott states at around 13:10

pure gold

Presentation slides: www.slideshare.net/SessionsEvents/3-ewa-dominowska-managing-machine-learning-projects-in-industry

Hey @MLconf, shouldn't you mention the title of the paper in the title rather than where the prof works? Titles can be edited, so it applies to the past uploads as well.

I had fun while learning about learning rate scheduling in this talk. Nice talk!

2:20 cracked me out - we all have been there

Results in 18:55

Very useful lecture, thanks for uploading such videos

@MLconf is the airbnb talk going to be shared ?

is the airbnb talk going to be shared ?

How to predict multiple replies from the single question?

when will we se the Uberstock below 25% . Race to the bottom.

ok, just seeing this now and there are likely even better videos now available although this one is terrific. One nit.... can you give the cost function in python. i don't see the code for that.