ICML
2006
IEEE
15 years 8 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...