keyword

unlabeled instances

Unlabeled instances are individual data points or examples within a dataset that contain observable input features but lack corresponding ground-truth target values, such as class labels or outcome categories. In machine learning, these instances represent raw observations that have not been tagged, categorized, or verified by human annotators or automated labeling processes. While standard supervised learning requires labeled examples to learn direct input-to-output mappings, unlabeled instances are widely utilized in unsupervised learning to discover underlying structure and patterns, in semi-supervised learning to augment limited labeled data, and in active learning to identify the most informative data points for targeted manual annotation.

1 item

Heterogeneous Uncertainty Sampling for Supervised Learning

David D. Lewis, Jason Catlett

OrganizationsAT&T Bell Laboratories

Why you should read this

Demonstrates that using a fast probabilistic classifier to actively select training instances for a more complex C4.5 rule induction model achieves lower error rates on text categorization tasks than random sampling sets ten times larger.

Uncertainty sampling methods iteratively request class labels for training instances whose classes are uncertain despite the previous labeled instances. These methods can greatly reduce the number of instances that an expert need label. One problem with this approach is that the classifier best suited for an application may be too expensive to train or use during the selection of instances. We test the use of one classifier (a highly efficient probabilistic one) to select examples for training another (the C4.5 rule induction program). Despite being chosen by this heterogeneous approach, the uncertainty samples yielded classifiers with lower error rates than random samples ten times larger.

Source

http://www.cs.cornell.edu/Courses/cs6740/2010fa/papers/lewis-catlett-uncertainty-sampling.pdf

Added

2026-09-25