Multiplicative Updates

Multiplicative weight update method

The multiplicative weights update method is an algorithmic technique used for decision making and prediction. It assigns initial weights to experts and updates them based on their performance, allowing the decision maker to iteratively decide which expert to follow. It has been discovered in many fields such as machine learning, optimization, theoretical computer science, and game theory.

1 courses cover this concept

CS 170: Efficient Algorithms and Intractable Problems

UC Berkeley

Spring 2020

This is an introductory course to computer science theory, exploring the design and analysis of various algorithms, number theory, and complexity. The prerequisites include familiarity with mathematical induction, big-O notation, basic data structures, and programming in a standard language.

No concepts data

+ 36 more concepts