We call a learning environment partially blind when there is an admixture of supervised and unsupervised (or blind) learning. Such situations typically arise in practice when supervised training data labelled by a teacher are scarce or expensive and are supplemented by inexpensive unlabelled (or blind) data available in relative profusion. Vapnik-Cervonenkis theory can be deployed in such settings to quantify the relative worth of supervision (and the lack thereof) in learning. We illustrate the nature of the tradeos possible in a simple setting of hyperplane decision functions and make explicit the role of dimensionality and side-information in these tradeos in the context of d-variate Gaussian mixtures.
|Journal||Proceedings - IEEE International Symposium on Circuits and Systems|
|State||Published - 1 Jan 2000|
|Event||Proceedings of the IEEE 2000 International Symposium on Circuits and Systems, ISCAS 2000 - Geneva, Switz, Switzerland|
Duration: 28 May 2000 → 31 May 2000
ASJC Scopus subject areas
- Electrical and Electronic Engineering