Why are non-ML people so fascinated with "unsupervised learning", which isn't even a thing? I blame Andrew Ng
@ferrouswheel you mean human-provided labels vs. machine-generated labels? The distinction is extremely specious
-
-
@fchollet can an autoencoder with a two node bottleneck learn face recognition? huge diff between answering a q vs finding an embedding. -
@kcimc@fchollet good embeddings support reasoning. below: exp psychology vector model of inductive reasoning (1983)pic.twitter.com/MPEUAE5mkx
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.