Check which algorithms could use the @reduction decorator from COMPSs, compare the performance, and incorporate it if it's better. Off the top of my head: KMeans, Gaussian Mixtures, LinearRegression, PCA, and probably more could use it. We should also check how to adapt the algorithms using an arity parameter.
Check which algorithms could use the @reduction decorator from COMPSs, compare the performance, and incorporate it if it's better.
Off the top of my head: KMeans, Gaussian Mixtures, LinearRegression, PCA, and probably more could use it.
We should also check how to adapt the algorithms using an arity parameter.