Heavy Hitters and Bernoulli Convolutions
Numerical Analysis
2019-05-29 v2 Machine Learning
Probability
Statistics Theory
Machine Learning
Statistics Theory
Abstract
A very simple event frequency approximation algorithm that is sensitive to event timeliness is suggested. The algorithm iteratively updates categorical click-distribution, producing (path of) a random walk on a standard -dimensional simplex. Under certain conditions, this random walk is self-similar and corresponds to a biased Bernoulli convolution. Algorithm evaluation naturally leads to estimation of moments of biased (finite and infinite) Bernoulli convolutions.
Cite
@article{arxiv.1905.08930,
title = {Heavy Hitters and Bernoulli Convolutions},
author = {Alexander Kushkuley},
journal= {arXiv preprint arXiv:1905.08930},
year = {2019}
}
Comments
1) fixed some typos and a reference 2) expanded section 3