We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.ML

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Machine Learning

Title: Improving image classifiers for small datasets by learning rate adaptations

Abstract: Our paper introduces an efficient combination of established techniques to improve classifier performance, in terms of accuracy and training time. We achieve two-fold to ten-fold speedup in nearing state of the art accuracy, over different model architectures, by dynamically tuning the learning rate. We find it especially beneficial in the case of a small dataset, where reliability of machine reasoning is lower. We validate our approach by comparing our method versus vanilla training on CIFAR-10. We also demonstrate its practical viability by implementing on an unbalanced corpus of diagnostic images.
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG); Applications (stat.AP)
Cite as: arXiv:1903.10726 [stat.ML]
  (or arXiv:1903.10726v2 [stat.ML] for this version)

Submission history

From: Sourav Mishra [view email]
[v1] Tue, 26 Mar 2019 08:22:01 GMT (467kb,D)
[v2] Wed, 27 Mar 2019 17:15:30 GMT (467kb,D)

Link back to: arXiv, form interface, contact.