Based in Sydney, Australia, Foundry is a blog by Rebecca Thao. Her posts explore modern architecture through photos and quotes by influential architects, engineers, and artists.

Episode 118 - Learning vs Memorization, Bitcoin Halvening, and Antibody Tests

Episode 118 - Learning vs Memorization, Bitcoin Halvening, and Antibody Tests

The Local Maximum cover a wide range of topics in this Local Maximum May news update with Max Sklar and cohost Aaron. Max said “this is one of the most fun episodes to do, but all the best stuff is at the end so you better listen through”.

Concept of the week is the separation of test and training set in machine learning. We discuss why it is important, how it separates learning and memorization, and where the grey areas are.

Other Topics include:
- The Bitcoin halving - a once-in-4-years watershed moment for the fledgling currency
- Revealing information from coronavirus antibody tests
- A 5 year old driver in Utah
- The predictive power of Demolition Man and other movies (at the end)

Links

NY Post: 5 Year old Caught Driving to California to Buy a Lamborghini

CBNC: Cuomo Shocked most new Coronavirus hospitalization are people staying home
QZ: New Coronavirus Antibody test is much more accurate
NYTimes: Coronavirus Antibody Prevalence in people who have had Covid-19

CointList on Medium: Is Bitcoin Designed to Rise Every Four Years (on the Halvening/Halving)
CoinMarketCap Halving Countdown (At time of recording, it was counting to the halving of 2020, but at time of writing, it is counting down to the halving of 2024)

Vulture.com: Interview with Demolition Man Writer Daniel Waters

Related Episodes

Episode 117 with Christian Hubbs about a dubious result that tested in the training set
Episode 16 on Overfitting and Under-fitting
Episode 1 on Bayesian Analysis of Rare Events

Episode 119 - Advocating for Bayesian Inference with Brian Blais

Episode 119 - Advocating for Bayesian Inference with Brian Blais

Episode 117 - A.I. Hype and the Expert Problem with Christian Hubbs

Episode 117 - A.I. Hype and the Expert Problem with Christian Hubbs