• Home
  • Weblog
  • Projects
  • Links
  • About

Why Momentum Really Works

We often think of Momentum as a means of dampening oscillations and speeding up the iterations, leading to faster convergence. But it has other interesting behavior. It allows a larger range of step-sizes to be used, and creates its own oscillations. What is going on?

April 05, 2017 optimization – gradient descent – momentum – visualization – distill.pub

Google Hash Code 2017 in a Knapsack

March 01, 2017 knapsack – hashcode – hackathon – combinatorics – optimization – google – caching – dynamic programming

A few days ago I participated in the 2017 Google Hash Code challenge, which was about to find an optimal configuration of videos to cache on different cache servers, given some requests and respective latencies from endpoints that may or may not be connected to individual cache servers. You can ...

Read more...
  • Social

    • GitHub
    • Keybase
  • Links

    • RSS (Blog)
    • RSS (Links)
    • Language Learning Toolkit
    • Batyr
  • GitHub

    Status updating...


© 2017 Markus Beuckelmann · Standing on the shoulders of giants: Pelican & Bootstrap (Theme based on pelican-bootstrap3) · Impressum

Back to top