One thought on “Gottingen talk

  1. Since i’m an interdisciplnary research scientist (with PhDs in Dilletantism, Procrastination and Discipl(ine)ship ) interested in education, I could relate to the discussion on ‘when learning fails’.

    Seems to boil down to 3 very short equations.

    The other equations in the paper look similar to ones I’ve seen before (in old papers by Cowan, Grossberg, Kuramoto , and statistical physics).

    I see WAT (Stone-Weirstrauss Approximation Theorem) makes a cameo appearence. One of my favorite theorems i don’t understand (though i call it piecewise linear approximation, or you can generalize to splines, wavelets, etc. There are maybe 10-20 theorems that would cover most of the phase space) .

    The ACM Turing award talk online was yesterday on ‘deep learning’ and was given by Yann Lecun (someone i had never heard of ) and likely may be related to this. (Some trace ‘deep learning’ back to to the wilson-cowan model; i trace it back to a popular article in Sci Am by John Hopfield on spin glasses and neural nets. ).

    (I was told about this by someone trained at U Toronto who actually took a course by LEH Trainor. retired from industry and has a recent paper I think that could be applied to learning and education if you change the word light to society. they say traffic flows follow a poisson distribution, which is a variant of gaussian.

    He also had seen the ACM Turing award talk by Judea Pearl on his own ‘causal reasoning’ approach— the differences that do/n’t make a difference. Pearl says deep learning is ‘just curve fitting’.

    Geoff Hinton previously of CMU shared this prize–he’s sort of connected to the ‘connectionist’ camp as opposed to innatist/chomskyian ‘ camp.

    I wonder if David Hilbert attended the Gottingen lecture–eg make a cameo appearance. .


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s