Tweetovi
- Tweetovi, trenutna stranica.
- Tweetovi i odgovori
- Medijski sadržaj
Blokirali ste korisnika/cu @_brohrer_
Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @_brohrer_
-
Prikvačeni tweet
Backpropagation is the beating heart of neural networks. Here's how it works. https://youtu.be/6BMwisTZFr4 pic.twitter.com/IpWTD2dglx
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Excellent context and a fascinating read from
@rabble.https://twitter.com/rabble/status/1224820389387223041 …Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
I'm teaching an Intro to Deep Learning workshop sequence at
@odsc East in April! Come join me for a full day of concepts and code, no prior experience required. https://odsc.com/speakers/introduction-to-deep-learning-neural-networks-i-concepts/ …https://odsc.com/speakers/introduction-to-deep-learning-neural-networks-ii-practice/ …Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Brandon Rohrer proslijedio/la je Tweet
The iRobot Comms team is hiring. Do you know someone in the Boston area with 1-2 years of comms experience who is excited about consumer tech, robotics, media relations, reviews and writing? I am proud to say that we have a great team.https://www.linkedin.com/jobs/view/1710976955/ …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Details of my implementation and some things I learned: https://e2eml.school/k_sparse_layer.html …
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
StatQuest by
@joshuastarmer is one of the best statistics and ML resources in existence. He's now committed to making tutorials for us full-time. StatQuest shows what educational internet can be in the right hands.https://statquest.org/video-index/Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Actual moral of the story: keep building, keep asking why and how. 11/11
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Most importantly, I understand so much more now than I would have if I had skipped right to the authors’ solution. Re-implementation gives you understanding. Re-derivation gives you a sense of ownership. 10/11
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Just kidding. My data was different than the original authors’ so my results were pretty different too. I did some small, but I think interesting, things differently. 9/11
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Moral of the story: everything interesting has already been done 8/11
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
I was thrilled by the result. I went to share it and a more careful search showed that the method is already been presented very thoroughly in a 2013 arxiv paper: k-Sparse Autoencoders by Alireza Makhzani and Brendan Frey https://arxiv.org/pdf/1312.5663.pdf … 7/11pic.twitter.com/7OH2UhZlT7
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
This problem is an instance of Sparse Dictionary Learning. https://en.wikipedia.org/wiki/Sparse_dictionary_learning … https://arxiv.org/pdf/1312.5663.pdf … 6/11
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
With sparsification, these nodes have much more structure and form a reasonable sparse basis. 5/11pic.twitter.com/30uQ1KU4gY
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Without sparsification, these nodes are information rich, but visually uninteresting. They look like static. 4/11pic.twitter.com/Q2t3pINI73
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Using it in an autoencoder lets you see what it does. You can visualize nodes in the bottleneck layer by finding what output image they produce if all the other nodes are zero. 3/11
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
So I made a layer that preserved the k node activities with the highest magnitudes. The rest it set to zero. https://gitlab.com/brohrer/cottonwood/-/blob/main/cottonwood/experimental/layers/sparsify.py … 2/11
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
A research story with a twist I looked for a neural network regularization method that limited the number of non-zero node activities (L0 group sparsity?) I couldn’t find one. 1/11
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
I figured out how to solve the problem of cheesy robot stock photos: Cheesy robot drawings
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Also http://e2eml.school/research (new ideas) http://e2eml.school/stickers (decorate your laptop) As always you are invited to fork my library repository (https://gitlab.com/brohrer/e2eml-library …) and use it as a template for your own blog.
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
I've included some handy shortcuts too http://e2eml.school/courses (course listing) http://e2eml.school/library (tutorial posts) http://e2eml.school/blog (Office Hours blog) http://e2eml.school/312 (Course 312. Find courses directly by number)
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.