Subsampling vs feature selection?

[nabenabe0928]

Since we often assume the manifold hypothesis or low-intrinsic dimensionality in most cases, I think it is better to reduce the dataset size by feature selection, i.e. the reduction over the column size, rather than subsampling, i.e. the reduction over the row size.

What do @ravinkohli think?

[ravinkohli]

For reference, autogluon has implemented feature pruning here