Papers › An Analysis of Unsupervised Pre-training in Light of Recent Advances

An Analysis of Unsupervised Pre-training in Light of Recent Advances

20 Dec 2014arXiv:1412.6597archive 2025-07-28

Tom Le Paine, Pooya Khorrami, Wei Han, Thomas S. Huang

Convolutional neural networks perform well on object recognition because of a number of recent advances: rectified linear units (ReLUs), data augmentation, dropout, and large labelled datasets. Unsupervised data has been proposed as another way to improve performance. Unfortunately, unsupervised pre-training is not used by state-of-the-art methods leading to the following question: Is unsupervised pre-training still useful given recent advances? If so, when? We answer this in three parts: we 1) develop an unsupervised method that incorporates ReLUs and recent unsupervised regularization techniques, 2) analyze the benefits of unsupervised pre-training compared to data augmentation and dropout on CIFAR-10 while varying the ratio of unsupervised to supervised samples, 3) verify our findings on STL-10. We discover unsupervised pre-training, as expected, helps when the ratio of unsupervised to supervised samples is high, and surprisingly, hurts when the ratio is low. We also use unsupervised pre-training with additional color augmentation to achieve near state-of-the-art performance on STL-10.

PaperPDFCode

Code

ifp-uiuc/an-analysis-of-unsupervised-pre-training-iclr-2015 officialmentioned in papermentioned on GitHub report
ifp-uiuc/anna officialmentioned in papermentioned on GitHub report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Data AugmentationImage ClassificationObject RecognitionUnsupervised Pre-training

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Classification CIFAR-10 An Analysis of Unsupervised Pre-training in Light of Recent Advances Percentage correct 86.7 #222 of 265 Archive leaderboard report
Image Classification STL-10 An Analysis of Unsupervised Pre-training in Light of Recent Advances Percentage correct 70.2 #91 of 117 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Dropout

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections