Title :
Estimation of Gaussian Mixtures from Moments
Author_Institution :
HNC Software / Fair Isaac Res., San Diego, CA, USA
Abstract :
We introduce a technique for estimating a Gaussian mixture model from estimates of the moments of a data set. This technique allows for model estimation without direct access to private data and without the need to store large amounts of data for online applications.
Keywords :
Gaussian distribution; data privacy; estimation theory; security of data; Gaussian mixture model estimation; data privacy; moment estimation; online application; Application software; Computer crime; Customer relationship management; Data privacy; Input variables; Law enforcement; Linear regression; Machine learning; Predictive models; Table lookup; gaussian mixture models; mixture estimation; online learning; unsupervised learning;
Conference_Titel :
Machine Learning and Applications, 2008. ICMLA '08. Seventh International Conference on
Conference_Location :
San Diego, CA
Print_ISBN :
978-0-7695-3495-4
DOI :
10.1109/ICMLA.2008.61
