https://github.com/conal/talk-2018-deep-learning-rebooted

"A Functional Reboot for Deep Learning", an invited talk for Summer BOB 2019 in Berlin
https://github.com/conal/talk-2018-deep-learning-rebooted

Last synced: about 2 months ago
JSON representation

"A Functional Reboot for Deep Learning", an invited talk for Summer BOB 2019 in Berlin

• Host: GitHub
• URL: https://github.com/conal/talk-2018-deep-learning-rebooted
• Owner: conal
• Created: 2018-01-06T23:56:42.000Z (about 7 years ago)
• Default Branch: master
• Last Pushed: 2019-11-11T20:03:33.000Z (about 5 years ago)
• Last Synced: 2024-08-02T10:27:34.777Z (6 months ago)
• Language: TeX
• Homepage:
• Size: 39.1 KB
• Stars: 52
• Watchers: 4
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: readme.md

Awesome Lists containing this project

• awesome-haskell-deep-learning - A Functional Reboot for Deep Learning (BOB 2019 Talk)

README

        
## A Functional Reboot for Deep Learning

Invited talk for [Summer BOB 2019 in Berlin](https://bobkonf.de/2019-summer/program.html).

You can find [the slides (PDF)](http://conal.net/talks/deep-learning-rebooted.pdf) in [my talks folder](http://conal.net/talks/). [A video (50 mins)](https://www.youtube.com/watch?v=Ns3DxUeCvRg) is available on YouTube.

**Abstract**:

 


In this talk, I want to begin a conversation about what is the essence of deep learning and how we can optimally support this essence in the form of a programming interface or language. I'll give you my own impressions, and I hope to provoke an ongoing conversation. Despite the phenomenal success of deep learning, it's my sense that most of the choices made in the theory and practice of deep learning are nonessential and even harmful (unnecessarily complex and limited). I'll suggest that a very small addition to a modern typed functional programming language such as Haskell yields an ideal basis for deep learning that is much simpler, more general, and more rigorous that currently popular approaches.