This lecture talks about the multiplicative update rule for learning linear classifiers. In particular, we will look at the Winnow algorithm and some variants.
Links and Resources
-
Avrim Blum’s notes on the algorithm
Further reading
-
Sanjeev Arora, Elad Hazan and Satyen Kale, The Multiplicative Weights Update Method: a Meta Algorithm and Applications
-
Nick Littlestone, Learning Quickly When Irrelevant Attributes Abound, /Machine Learning/ 1988.
-
Jyrki Kivinen, Manfred Warmuth and Peter Auer, The Perceptron algorithm versus Winnow: linear versus logarithmic mistake bounds when few input variables are relevant, /Artificial Intelligence/, 1997.