ICML
2006
IEEE
15 years 11 days ago
2006
IEEE
In machine learning problems with tens of thousands of features and only dozens or hundreds of independent training examples, dimensionality reduction is essential for good learni...