Strawberry picking now needs me every day, as opposed to less days!
But Alisi and Joe are letting me just work half-days because I'm
writing a book.
My chapter one is short, but kicks off with a very dense common lisp
function that spawns hopfield net closures. The dense code serves a
few purposes. Latently, the closures are useable and useful in common
lisp (and via ecl, C). This code is also my own definition of a
binary Hopfield network with rectified quadratic activation,
connected from Krotov and Hopfield 2016. However, the logic and code
itself is not very desirable to play with changing.
Anyway, I did the deep-learning-doing-XOR thing. Participation and
criticism requested!
! I didn't write the appendix yet, wherein I intend to explain asdf
(basically put my cl-binry-hop/ gopher directory into ~/common-lisp/
in order to use it like chapter 1).