Loading [a11y]/accessibility-menu.js
Progressive Stochastic Learning for Noisy Labels | IEEE Journals & Magazine | IEEE Xplore

Progressive Stochastic Learning for Noisy Labels


Abstract:

Large-scale learning problems require a plethora of labels that can be efficiently collected from crowdsourcing services at low cost. However, labels annotated by crowdso...Show More

Abstract:

Large-scale learning problems require a plethora of labels that can be efficiently collected from crowdsourcing services at low cost. However, labels annotated by crowdsourced workers are often noisy, which inevitably degrades the performance of large-scale optimizations including the prevalent stochastic gradient descent (SGD). Specifically, these noisy labels adversely affect updates of the primal variable in conventional SGD. To solve this challenge, we propose a robust SGD mechanism called progressive stochastic learning (POSTAL), which naturally integrates the learning regime of curriculum learning (CL) with the update process of vanilla SGD. Our inspiration comes from the progressive learning process of CL, namely learning from “easy” tasks to “complex” tasks. Through the robust learning process of CL, POSTAL aims to yield robust updates of the primal variable on an ordered label sequence, namely, from “reliable” labels to “noisy” labels. To realize POSTAL mechanism, we design a cluster of “screening losses,” which sorts all labels from the reliable region to the noisy region. To sum up, POSTAL using screening losses ensures robust updates of the primal variable on reliable labels first, then on noisy labels incrementally until convergence. In theory, we derive the convergence rate of POSTAL realized by screening losses. Meanwhile, we provide the robustness analysis of representative screening losses. Experimental results on UCI1 simulated and Amazon Mechanical Turk crowdsourcing data sets show that the POSTAL using screening losses is more effective and robust than several existing baselines.1UCI is the abbreviation of University of California Irvine.
Published in: IEEE Transactions on Neural Networks and Learning Systems ( Volume: 29, Issue: 10, October 2018)
Page(s): 5136 - 5148
Date of Publication: 05 February 2018

ISSN Information:

PubMed ID: 29994430

Funding Agency:


References

References is not available for this document.