Closed rob-nyu closed 10 years ago
Figure out how to implement PCA by testing on the output of the boilerplate tf-idf model. May be used for huge HTML spaces.
Done. Memory may be an issue though. I could scale 92xxx features to 100, but if I try to scale down to 1000, my poor laptop runs out of memory
Figure out how to implement PCA by testing on the output of the boilerplate tf-idf model. May be used for huge HTML spaces.