←back to thread

294 points imurray | 2 comments | | HN request time: 0s | source
Show context
srean ◴[] No.44066091[source]
The problem of predicting tides was so important that it attracted many Physics and Maths heavy weights. You can well imagine how important predicting tides would have been for D-day landing.

One related fascinating historical artifact is the special purpose analogue computer designed by Lord Kelvin in the 1860s based on Fourier series, harmonic analysis. Think difference engine in it's cogs and cams glory, but special purpose.

https://en.m.wikipedia.org/wiki/Tide-predicting_machine

Possibly one of the first examples of Machine learning, with Machine in capital 'M'. It incorporated recent tidal observations to update it's prediction.

Note that sinusoids are universal approximators for a large class of functions, an honour that is by no means restricted to deep neural nets.

George Darwin (Charles Darwin's son) was a significant contributor in the design and upgrade of the machine.

https://en.m.wikipedia.org/wiki/George_Darwin

Other recognizable names who worked on tide prediction problem were Thomas Young (of double slit experiment fame) and Sir George Airy (of Airy disk fame).

replies(4): >>44066143 #>>44066655 #>>44069709 #>>44070084 #
CGMthrowaway ◴[] No.44066655[source]
Have you seen the SF bay model? https://www.youtube.com/watch?v=i70wkxmumAw
replies(2): >>44066710 #>>44069736 #
1. srean ◴[] No.44066710[source]
That was so fascinating. Thank you.
replies(1): >>44067461 #
2. synalx ◴[] No.44067461[source]
If you're ever in SF, it's really worth going to see. Such a cool mixture of art and technology.