Class meeting for 10-605 Parallel Perceptrons 2

This is one of the class meetings on the schedule for the course Machine Learning with Large Datasets 10-605 in Fall_2016.

Slides

Perceptrons, continued:

• Slides in Powerpoint
• Slides in PDF

Parallel perceptrons with iterative parameter mixing:

• Slides in Powerpoint
• Slides in PDF

Readings for the Class

• Distributed Training Strategies for the Structured Perceptron, R. McDonald, K. Hall and G. Mann, North American Association for Computational Linguistics (NAACL), 2010.

Optional Readings

• Notes on voted perceptron.

What you should remember

• The averaged perceptron and the voted perceptron
• Approaches to parallelizing perceptrons (and other on-line learning methods, like SGD)
  • Parameter mixing
  • Iterative parameter mixing (IPM)
• The meaning and implications of the theorems given for convergence of the basic perceptron and the IPM version