170
click to vote
DAGM
2011
Springer
14 years 3 months ago
2011
Springer
Many unsupervised learning algorithms make use of kernels that rely on the Euclidean distance between two samples. However, the Euclidean distance is optimal for Gaussian distribut...