Comments for Neural Networks Blog

Comment on Unsupervised Learning by Mdemon

Mdemon — Fri, 16 Sep 2022 14:56:17 +0000

I love Free Fire

Comment on Unsupervised Learning by TechieTrends (@trendstechie)

TechieTrends (@trendstechie) — Thu, 04 Oct 2018 07:35:20 +0000

Great and interesting article. Nice information. Kindly know network power solutions

Comment on Universal Approximation and Depth by Filip

Filip — Fri, 25 Aug 2017 17:28:10 +0000

ilyasu1, you say in the second comment that there are functions that deep networks cannot represent at all. Could you give some example o explanation of why? Tanks

Comment on Unsupervised Learning by Unsupervised Learning – The Conscience of A Libertarian

Unsupervised Learning – The Conscience of A Libertarian — Sun, 31 Jan 2016 22:40:51 +0000

[…] Originally here by Ilya Sutskever September 8, 2013 – 3:17 pm […]

Comment on The Miracle of the Boltzmann Machine by 1 – The Miracle of the Boltzmann Machine – blog.offeryour.com

1 – The Miracle of the Boltzmann Machine – blog.offeryour.com — Sun, 25 Jan 2015 01:52:48 +0000

[…] Read Startup article here: https://theneural.wordpress.com/2011/07/08/the-miracle-of-the-boltzmann-machine/ […]

Comment on The Miracle of the Boltzmann Machine by 1 – The Miracle of the Boltzmann Machine – Exploding Ads

1 – The Miracle of the Boltzmann Machine – Exploding Ads — Sun, 25 Jan 2015 01:50:18 +0000

[…] : https://theneural.wordpress.com/2011/07/08/the-miracle-of-the-boltzmann-machine/ […]

Comment on The useless beauty of Reinforce by timvieira

timvieira — Sun, 21 Dec 2014 02:06:37 +0000

I love the title of this post.

Comment on The futility of gigantic training sets with simple models by robertoschler

robertoschler — Mon, 08 Dec 2014 00:15:56 +0000

Great article! Minor typo: “better at trainign deep neural networks” -> training

Comment on Unsupervised Learning by Rob Lang

Rob Lang — Mon, 03 Nov 2014 11:26:49 +0000

Unsupervised learning is being used in taxonomic systems where human bias means that the classification of species is warped by ego. Unsupervised learning can also be used to detect novelty in non-stationary environments. Unsupervised learning is best used for exploring a dataset, which doesn’t have a natural industrial fit because it’s exploring rather than solving.

Comment on Universal Approximation and Depth by Deep Learning oral traditions - O'Reilly Radar

Deep Learning oral traditions - O'Reilly Radar — Mon, 28 Jul 2014 19:24:34 +0000

[…] on most machine-learning tasks. However this universal approximation property came at a steep cost: the requisite (single hidden layer) neural networks were exponentially inefficient to construct (you needed a neuron for every possible input). For a while neural networks took a backseat to more […]