align-toparrow-leftarrow-rightbackbellblockcalendarcamerachatcheckchevron-downchevron-leftchevron-rightchevron-small-downchevron-small-leftchevron-small-rightchevron-small-upchevron-upcircle-with-crosscrosseditfacebookglobegoogleimagesinstagramlocation-pinmagnifying-glassmailmoremuplabelShape 3 + Rectangle 1outlookpersonplusImported LayersImported LayersImported Layersshieldstartwitteryahoo

Future salon: machine learning

Topics covered will include beating experts at wine tasting,
predicting housing prices, predicting which products from a chip
manufacturing process will pass testing, how to do optical character
recognition (OCR), how to classify emails as spam or not spam, how to
reduce color palettes in images, how to detect anomalies in a
manufacturing dataset, and how to make a movie recommendation system. Based on the Coursera/Stanford course by Andrew Ng (who used machine learning to teach model helicopters to fly stunts, but that will not be covered ;) .

Join or login to comment.

  • Brian K.

    FYI: Got delayed with a work thing. Might not make it or just late.

    August 18, 2013

  • David V.

    Very interested in this, I have designed the hardware for some pretty intense ML systems but the software was done by a team of PHD students behind a curtain. :( I'm wondering how much of this can be applied to systems with only microcontrollers and limited ram.

    August 12, 2013

    • Wayne R.

      The math will work on limited systems, but you might have some trouble getting a lot of data in & out. (Learning algorithms like having a lot of data). The other thing you could do is do the learning on a big system, but once you're done transfer the resulting parameters to the limited system that uses them for something.

      August 12, 2013

    • David V.

      As far as I've gone with this, was to take three digital sensors (pulse width info) and four analog sensors, all sampled at rates appropriate to the sensors but in the 1-4 kS/S range, and calculate shannon's entropy on them, then display that on an LCD graph. That was on an AVR Mega644. The fun part was reworking the entropy formula so that I could get it all with integer math. But about that point the project was defunded. :( The PHDs were doing recognition of people in data from four imagers coming in at 27MP/s on four channels, into an FPGA. I didn't get to see any of how that worked. My end was about sensing vehicles by their magnetic and microwave doppler signatures.

      August 12, 2013

7 went

Our Sponsors

People in this
Meetup are also in:

Sign up

Meetup members, Log in

By clicking "Sign up" or "Sign up using Facebook", you confirm that you accept our Terms of Service & Privacy Policy