Loading Now

Summary of Smoothed Online Classification Can Be Harder Than Batch Classification, by Vinod Raman et al.


Smoothed Online Classification can be Harder than Batch Classification

by Vinod Raman, Unique Subedi, Ambuj Tewari

First submitted to arxiv on: 24 May 2024

Categories

  • Main: Machine Learning (cs.LG)
  • Secondary: None

     Abstract of paper      PDF of paper


GrooveSquid.com Paper Summaries

GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!

Summary difficulty Written by Summary
High Paper authors High Difficulty Summary
Read the original abstract here
Medium GrooveSquid.com (original content) Medium Difficulty Summary
This paper investigates online classification under smoothed adversaries, where an adversary draws examples from a distribution with bounded density relative to a known base measure. The authors build upon previous works showing that smoothed online learning is equivalent to batch learning in certain settings, but they also show that unbounded label spaces can make smoothed online classification harder than iid batch classification. They construct a hypothesis class that is learnable in the batch setting but not in the smoothed online setting, highlighting a condition for PAC learnability being sufficient for smoothed online learnability.
Low GrooveSquid.com (original content) Low Difficulty Summary
This paper looks at how to classify things online when there’s an “adversary” trying to trick you. The adversary draws examples from a special kind of distribution that has limits. The authors show that this can be tricky, especially if there are lots of different labels. They create an example where something that works well in one situation doesn’t work as well in another.

Keywords

» Artificial intelligence  » Classification  » Online learning