Message from JavaScript discussions

December 2017

β€” Never knew thatπŸ‘πŸ‘πŸ‘


Digital version does math with functional units of a processor or some other special hardware, analog version does math with direct current electriciy, meaning each neuron simply raises or lowers voltage in some way

β€” One is virtual (usually) while the other is physical

β€” So how does 3d takes various forms if datasets

β€” TensorFlow for example simulates a virtual tensor field, while an AANN would be a real life tensor field that behaves the same but operates on very different principles

Message permanent page

β€” Now I get it

β€” Should operate the same as the virtual one

β€” How exactly

β€” Actually my goal is to turn the virtual tensor fields into the physical one, by translating it with software

Message permanent page

β€” Let's say it binary classification

β€” How do compute that

β€” Good luck with that