An Improved Exponential-Time Approximation Algorithm for Fully-Alternating Games Against Nature

Explore an advanced algorithm for fully-alternating games against nature in this 33-minute IEEE conference talk by Andrew Drucker from the University of Chicago at FOCS 2020. Delve into topics such as adversarial versions, the "MA" formula perspective, and composition of algorithms. Learn about algorithms with baseline advice, composition and defect, "steepener" algorithms, and their actions. Examine decrements and decisiveness before reaching the concluding remarks on this improved exponential-time approximation approach.

Intro
Adversarial version
Games against Nature
"MA" formula perspective
Basic ideas
Composition of algorithms
Algorithms with baseline advice
Composition and defect
"Steepener" algorithms
Steepener action
Decrements and decisiveness
Conclusions