Local Linear Discriminant Analysis (LLDA)

A local version of LDA. This algorithm performs with the same strategy as LDA, but it makes use of an affinity matrix, that is, a sparse matrix that measures the closeness between the samples, and that defines the local structure to be learned.

Watch the full LLDA documentation here.

References

Masashi Sugiyama “Dimensionality reduction of multimodal labeled data by local fisher discriminant analysis”. In: Journal of Machine Learning Research, 2007, vol 8, May, pages 1027-1061.