New tutorial: building autoencoders in Keras. http://blog.keras.io/building-autoencoders-in-keras.html …pic.twitter.com/ehYr5cGutN
-
minor comment: t-SNE usually preceded 30+ dim PCA on data, only "genuinely" high rank data would profit from something like AE
Thanks! PCA + t-SNE is a combo I see frequently, but as I understand a multilayer AE can do a better job on high-dim datasets...
-
-
i want to believe! but in my experience improving on PCA + t-SNE is very difficult. here's a comparison:https://github.com/kylemcdonald/t-SNEPreprocessingComparison/blob/master/t-SNE%20Preprocessing%20Comparison.ipynb …
-
@github t-SNE is amazing but it's far slower and more memory hungry than auto encoders & can't online transform new points.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.