Least squares linear discriminant analysis

By Jieping Ye. International Conference on Machine Learning, 2007

309 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Foundations for Envelope Models and Methods (2015, methodology): “It has been shown (Ye 2007) that two-class LDA is equivalent to a least squares problem: no matter how we code Y for two classes, the least squares fitted coefficient vector β̂OLS is always proportional to the LDA direction.”
Hypergraph spectral learning for multi-label classification (2008, background): “For multi-class or multi-label problems, a similar definition can be applied for each label separately [12, 24].”
Distribution Calibration in Riemannian Symmetric Space (2011, background): “I. INTRODUCTION CROSS-DOMAIN learning algorithms leverage knowl-edge learned from the source domain for use in the target domain, where both domains are different but related [10].”
Maximum Likelihood Decision Fusion for Weapon Classification in Wireless Acoustic Sensor Networks (2017, methodology): “LS-LDA is a linear classification method, albeit it is possible to approximate nonlinear boundaries with a piece-wise linear function by assigning different classifiers to different regions of the feature space as the tree does.”
Spectral Embedded Clustering: A Framework for In-Sample and Out-of-Sample Spectral Clustering (2011, methodology): “On the other hand, our proposed SEC framework relies on the linear property of the cluster assignment matrix, which is usually satisfied for real-world data [34].”