Learning with CVaR-based feedback under potentially heavy tails

• URL: http://arxiv.org/abs/2006.02001v1
• Date: Wed, 3 Jun 2020 01:08:29 GMT
• Title: Learning with CVaR-based feedback under potentially heavy tails
• Authors: Matthew J. Holland, El Mehdi Haress
• Abstract summary: We study learning algorithms that seek to minimize the conditional value-at-risk (CVaR) We first study a general-purpose estimator of CVaR for potentially heavy-tailed random variables. We then derive a new learning algorithm which robustly chooses among candidates produced by gradient-driven sub-processes.
• Score: 8.572654816871873
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We study learning algorithms that seek to minimize the conditional value-at-risk (CVaR), when all the learner knows is that the losses incurred may be heavy-tailed. We begin by studying a general-purpose estimator of CVaR for potentially heavy-tailed random variables, which is easy to implement in practice, and requires nothing more than finite variance and a distribution function that does not change too fast or slow around just the quantile of interest. With this estimator in hand, we then derive a new learning algorithm which robustly chooses among candidates produced by stochastic gradient-driven sub-processes. For this procedure we provide high-probability excess CVaR bounds, and to complement the theory we conduct empirical tests of the underlying CVaR estimator and the learning algorithm derived from it.

Related papers

• Provably Efficient CVaR RL in Low-rank MDPs [58.58570425202862]
  We study risk-sensitive Reinforcement Learning (RL) We propose a novel Upper Confidence Bound (UCB) bonus-driven algorithm to balance interplay between exploration, exploitation, and representation learning in CVaR RL. We prove that our algorithm achieves a sample complexity of $epsilon$-optimal CVaR, where $H$ is the length of each episode, $A$ is the capacity of action space, and $d$ is the dimension of representations.
  arXiv  Detail & Related papers  (2023-11-20T17:44:40Z)
• Towards Diverse Evaluation of Class Incremental Learning: A Representation Learning Perspective [67.45111837188685]
  Class incremental learning (CIL) algorithms aim to continually learn new object classes from incrementally arriving data. We experimentally analyze neural network models trained by CIL algorithms using various evaluation protocols in representation learning.
  arXiv  Detail & Related papers  (2022-06-16T11:44:11Z)
• Distributionally Robust Models with Parametric Likelihood Ratios [123.05074253513935]
  Three simple ideas allow us to train models with DRO using a broader class of parametric likelihood ratios. We find that models trained with the resulting parametric adversaries are consistently more robust to subpopulation shifts when compared to other DRO approaches.
  arXiv  Detail & Related papers  (2022-04-13T12:43:12Z)
• Boosted CVaR Classification [44.731468512484824]
  A popular approach to maximize the model's tail performance is to minimize the CVaR loss. We show that for classification tasks where models are evaluated by the zero-one loss, the minimizer of the average zero-one loss also minimizes the CVaR zero-one loss. We propose the Boosted CVaR Classification framework which is motivated by a direct relationship between CVaR and a classical boosting algorithm called LPBoost.
  arXiv  Detail & Related papers  (2021-10-26T18:27:25Z)
• Risk Minimization from Adaptively Collected Data: Guarantees for Supervised and Policy Learning [57.88785630755165]
  Empirical risk minimization (ERM) is the workhorse of machine learning, but its model-agnostic guarantees can fail when we use adaptively collected data. We study a generic importance sampling weighted ERM algorithm for using adaptively collected data to minimize the average of a loss function over a hypothesis class. For policy learning, we provide rate-optimal regret guarantees that close an open gap in the existing literature whenever exploration decays to zero.
  arXiv  Detail & Related papers  (2021-06-03T09:50:13Z)
• Risk Bounds and Rademacher Complexity in Batch Reinforcement Learning [36.015585972493575]
  This paper considers batch Reinforcement Learning (RL) with general value function approximation. The excess risk of Empirical Risk Minimizer (ERM) is bounded by the Rademacher complexity of the function class. Fast statistical rates can be achieved by using tools of local Rademacher complexity.
  arXiv  Detail & Related papers  (2021-03-25T14:45:29Z)
• Unbiased Risk Estimators Can Mislead: A Case Study of Learning with Complementary Labels [92.98756432746482]
  We study a weakly supervised problem called learning with complementary labels. We show that the quality of gradient estimation matters more in risk minimization. We propose a novel surrogate complementary loss(SCL) framework that trades zero bias with reduced variance.
  arXiv  Detail & Related papers  (2020-07-05T04:19:37Z)
• Reparameterized Variational Divergence Minimization for Stable Imitation [57.06909373038396]
  We study the extent to which variations in the choice of probabilistic divergence may yield more performant ILO algorithms. We contribute a re parameterization trick for adversarial imitation learning to alleviate the challenges of the promising $f$-divergence minimization framework. Empirically, we demonstrate that our design choices allow for ILO algorithms that outperform baseline approaches and more closely match expert performance in low-dimensional continuous-control tasks.
  arXiv  Detail & Related papers  (2020-06-18T19:04:09Z)
• Information-theoretic analysis for transfer learning [5.081241420920605]
  We give an information-theoretic analysis on the generalization error and the excess risk of transfer learning algorithms. Our results suggest, perhaps as expected, that the Kullback-Leibler divergence $D(mu||mu')$ plays an important role in characterizing the generalization error.
  arXiv  Detail & Related papers  (2020-05-18T13:23:20Z)
• Statistical Learning with Conditional Value at Risk [35.4968603057034]
  We propose a risk-averse statistical learning framework wherein the performance of a learning algorithm is evaluated by the conditional value-at-risk (CVaR) of losses rather than the expected loss.
  arXiv  Detail & Related papers  (2020-02-14T00:58:34Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.