New research explores online learning for score-driven filters

A new research paper published on arXiv details advancements in online learning for score-driven filters. The study focuses on optimizing the gain parameter, which controls the update magnitude in these filters, by treating it as a decision variable. Researchers developed a method using a Kullback-Leibler objective and mirror-descent geometries to establish dynamic-regret bounds, showing improved performance in simulations and practical applications like predicting equity-index volatilities. AI

IMPACT This research could lead to more adaptive and accurate predictive models in financial markets and other time-series analysis applications.

RANK_REASON The cluster contains a research paper published on arXiv detailing a new method for online learning in score-driven filters. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv stat.ML →

• Agas
• alphaXiv
• Bregman
• CatalyzeX
• DagsHub
• Gotit.pub
• Hugging Face
• IArxiv
• Kullback--Leibler divergence
• mirror
• ScienceCast

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →