[LOCKSUIT]

Do ANNs just do these 4 things? Small brainstorm - help

Text/vision/sound nets just seem to do these 4 things:
1) Prediction/expectation/superresolution/expansion
2) Translation/recognition (I see a cat! I also see a dog in that cat!)
3) Summarization/shrink to lower quality/quantity
4) Segmentation/parse where or when is it parts go and re-/connect in a sequence

If you think about it, these constitute this:
1) What comes with What
2) What is What.
3) What goes Where
--- Prediction, Recognition, Segmentation.

So 3 golden things then, not 4. Because summarize=predict/expansion (What comes with What).

IOW, tell me some applications ANNs do. I bet they fall under those 3 things.

[HS]

Yeah, I tried looking up applications but that seems to cover it. Sometimes used in combination like when playing board games, but it's like the RGB of useful thought.

[LOCKSUIT]

K clustering and algorithms should all also fall under my rules.

clustering = What is What !

No surprise it is used in relational clusters in high dimensional space in neural networks used for Translation queries. Like word2vec.