Reproduce “Self-normalized Bernstein Inequalities for Regenerative Processes and Applications to Learning Without Independence” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Imon Banerjee, Pierre Alquier

The authors derive a fixed-horizon Bernstein bound for bounded-loss averages on regenerative processes, using observed regeneration-cycle lengths and completed-cycle losses to adapt to dependence. Marked cycles are assumed i.i.d., and the bound retains a nonempirical cycle-tail remainder. Under stated moment and initialization conditions, the minimax estimation radius scales as √(m/n), for integer mean cycle length m.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.