Adaptive Classification Using Incremental Linearized Kernel Embedding

Hall, John J.; Robbiano, Christopher; Azimi-Sadjadi, M.R.

doi:10.1109/tsp.2022.3162407

Cited by 1 publication

(1 citation statement)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…'expert advice') in real-time, for various prediction tasks in numerous fields of science and engineering. Some examples of prediction tasks pertaining to signal processing and system automation include: adaptive filtering [2], [3], predictive control [4], feature extraction [5], [6], parameter estimation [7], [8], non-convex optimization [9], [10], classification [11]- [14], time-series analysis [15], speech de-reverberation [16].…”

Section: Introductionmentioning

confidence: 99%

Optimal Tracking in Prediction with Expert Advice

Gokcesu¹,

Kozat²

2022

Preprint

View full text Add to dashboard Cite

We study the prediction with expert advice setting, where the aim is to produce a decision by combining the decisions generated by a set of experts, e.g., independently running algorithms. We achieve the min-max optimal dynamic regret under the prediction with expert advice setting, i.e., we can compete against time-varying (not necessarily fixed) combinations of expert decisions in an optimal manner. Our end-algorithm is truly online with no prior information, such as the time horizon or loss ranges, which are commonly used by different algorithms in the literature. Both our regret guarantees and the min-max lower bounds are derived with the general consideration that the expert losses can have time-varying properties and are possibly unbounded. Our algorithm can be adapted for restrictive scenarios regarding both loss feedback and decision making. Our guarantees are universal, i.e., our end-algorithm can provide regret guarantee against any competitor sequence in a min-max optimal manner with logarithmic complexity. Note that, to our knowledge, for the prediction with expert advice problem, our algorithms are the first to produce such universally optimal, adaptive and truly online guarantees with no prior knowledge.This study is partially supported by Turkcell Technology within the framework of 5G and Beyond Joint Graduate Support Programme coordinated by Information and Communication Technologies Authority.

show abstract