Aaditya Ramdas (CMU) -- Exponential line-crossing inequalities

preview_player
Показать описание
MIFODS - Stochastics and Statistics Seminar.
Cambridge, US
April 12, 2019
  1. FODSI
Рекомендации по теме
Aaditya Ramdas (CMU) 0:54:26

Aaditya Ramdas (CMU) -- Exponential line-crossing inequalities

Aaditya Ramdas on 0:57:54

Aaditya Ramdas on Exponential line-crossing inequalities

[AISTATS 2023 Oral] 0:11:32

[AISTATS 2023 Oral] Huber-Robust Confidence Sequences --- Hongjian Wang & Aaditya Ramdas (CMU)

'The Numeraire E-Variable 1:08:20

'The Numeraire E-Variable and Reverse Information Projection' by Aaditya Ramdas

Aaditya Ramdas: Comparing 0:59:53

Aaditya Ramdas: Comparing sequential forecasters

Aaditya Ramdas - 1:20:08

Aaditya Ramdas - Betting scores, e-values and martingales

AI4OPT Tutorial Lectures: 1:46:36

AI4OPT Tutorial Lectures: A Martingale Theory of Evidence (Part I)

An Analysis of 0:24:04

An Analysis of Active Learning with Uniform Feature Noise -- Aaditya Ramdas

Dynamic Algorithms for 0:13:10

Dynamic Algorithms for Online Multiple Testing

recitation 1 0:49:58

recitation 1

Aaditya Ramdas: Universal 0:34:31

Aaditya Ramdas: Universal inference using the split likelihood ratio test

recitation 9 0:58:15

recitation 9

Assumption-free uncertainty quantification 0:15:13

Assumption-free uncertainty quantification for ML

Weekly Research Seminar 1:01:55

Weekly Research Seminar with Prof. Aaditya Ramdas - Distribution-free uncertainty quantification

Statistics and Data 0:01:57

Statistics and Data Science Major Spotlight Video

Confidence sequences for 0:48:52

Confidence sequences for sampling without replacement

lecture 18: semidefinite 1:11:52

lecture 18: semidefinite programming

Asymptotic behavior of 0:09:52

Asymptotic behavior of $\ell_q$-based Laplacian regularization in semi-supervised learning

Lecture 14: Boosting 1:17:05

Lecture 14: Boosting

AI4OPT Tutorial Lectures: 1:42:44

AI4OPT Tutorial Lectures: A Martingale Theory of Evidence (Part II)

Sequential Estimation of 1:12:53

Sequential Estimation of Quantiles with Applications to A/B-testing and Best-arm Identification

Merging Uncertainty Sets 0:56:11

Merging Uncertainty Sets via Majority Vote

Power of Active 0:51:35

Power of Active Sampling for Unsupervised Learning

DSI: Aarti Singh 1:31:28

DSI: Aarti Singh (CMU) – Learning and Decision Making with Preferential Supervision

Комментарии
Автор

Xj Chen




In the paper of ``Exponential Line Crossing Inequalities'', the authors are using the idea of sub \psi and variance process V_t to unify all kinds of concentration inequalities. Actually, in Chen's paper, the same idea was proposed by using the function \varphi and variance bounding function V_t to unify a wide range of concentration inequalities.

xjchen