Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Impossible Tuning Made Possible: A New Expert Algorithm and Its Applications

About

We resolve the long-standing "impossible tuning" issue for the classic expert problem and show that, it is in fact possible to achieve regret $O\left(\sqrt{(\ln d)\sum_t \ell_{t,i}^2}\right)$ simultaneously for all expert $i$ in a $T$-round $d$-expert problem where $\ell_{t,i}$ is the loss for expert $i$ in round $t$. Our algorithm is based on the Mirror Descent framework with a correction term and a weighted entropy regularizer. While natural, the algorithm has not been studied before and requires a careful analysis. We also generalize the bound to $O\left(\sqrt{(\ln d)\sum_t (\ell_{t,i}-m_{t,i})^2}\right)$ for any prediction vector $m_t$ that the learner receives, and recover or improve many existing results by choosing different $m_t$. Furthermore, we use the same framework to create a master algorithm that combines a set of base algorithms and learns the best one with little overhead. The new guarantee of our master allows us to derive many new results for both the expert problem and more generally Online Linear Optimization.

Liyu Chen, Haipeng Luo, Chen-Yu Wei• 2021

Related benchmarks

TaskDatasetResultRank
Online Model SelectionRotated MNIST Abrupt Drift
Cumulative Loss158.1
10
ClassificationAirlines (test)
Normalized Cumulative Loss100
5
ClassificationPowersupply (test)
Normalized Cumulative Loss100
5
ClassificationarXiv (test)
Normalized Cumulative Loss100
5
ClassificationForest (test)
Normalized Cumulative Loss100
5
ClassificationHuffPost (test)
Normalized Cumulative Loss100
5
ClassificationRialto (test)
Normalized Cumulative Loss100
5
ClassificationWeather (test)
Normalized Cumulative Loss100
5
Online Model SelectionRotated MNIST Incremental Drift
Cumulative Loss282.1
5
Regressionbikesharing (test)
Normalized Cumulative Loss100
5
Showing 10 of 11 rows

Other info

Follow for update