Foro1GH's profile picture. I model systems for a living, and live for the mystery beyond them.

Nana Quame

@Foro1GH

I model systems for a living, and live for the mystery beyond them.

. @OsagyefoNKB @HubertBaid80747 The "12 Pillars" manifesto is ambitious but not a policy document. To fairly evaluate feasibility, we need an actionable framework with timelines, funding sources, and implementation plans. I’ve combed your websites for such details to no avail.…


Nana Quame repostou

To all those who sent me the article claiming to rebut my essay on why Ghana Card is a ripoff, here is my reaction. The author of that article has a whole lot of explaining to do if his views are to be taken seriously. #SaveGhanaCard brightsimons.com/2024/04/02/the…


Nana Quame repostou

Bayesian nugget of the day: Metropolis-Hastings. 1) Draw a candidate θ from a proposal q(θ|θₜ), 2) Accept θ if u < π(θ)/π(θₜ) q(θₜ|θ)/q(θ|θₜ) for u~U[0,1], 3) Repeat. Beautifully, this results in a Markov chain whose stationary distribution coincides with the target π!


Nana Quame repostou

Here's what the future of education looks like: 1. Teaching will become an extremely lucrative profession. Salaries will follow a power law. The best teachers will make millions of dollars per year and teach thousands of students every year. In fact, this is already happening.


Nana Quame repostou

A non-linearity that works much better than ReLUs. The work described in this video might also be relevant to understanding grid cells. youtube.com/watch?v=Q2fLWG…

geoffreyhinton's tweet card. Implicit Neural Representations with Periodic Activation Functions

youtube.com

YouTube

Implicit Neural Representations with Periodic Activation Functions


Nana Quame repostou

#Rstats fiends and Bayesian freaks, it’s finally here. The inaugural 0.0.1 version of my translation of the second edition of @rlmcelreath’s “Statistical rethinking” into #brms and the #tidyverse is up! bookdown.org/content/4857/ 1/5


Nana Quame repostou

Like everyone else I've had many things on my mind this last week, most of them sad. On a positive note, I'd like to share the story of one of my favorite mathematicians: David Blackwell. His work had a huge impact on me, and his story is not shared often enough.

ChrKroer's tweet image. Like everyone else I&apos;ve had many things on my mind this last week, most of them sad. On a positive note, I&apos;d like to share the story of one of my favorite mathematicians: David Blackwell. 

His work had a huge impact on me, and his story is not shared often enough.

Nana Quame repostou

Tomorrow's ISSI speaker is Małgorzata Bogdan, who will speak about an adaptive Bayesian version of the SLOPE algorithm, for regression with missing covariate values. Discussion by Cynthia Rush to follow. Join at: sites.google.com/view/selective…

wfithian's tweet image. Tomorrow&apos;s ISSI speaker is Małgorzata Bogdan, who will speak about an adaptive Bayesian version of the SLOPE algorithm, for regression with missing covariate values. Discussion by Cynthia Rush to follow. 

Join at: sites.google.com/view/selective…

Nana Quame repostou

Bayesian continuous-time Kaplan-Meier survival function fit with #brms and plotted in a stepped line/ribbon plot with #tidybayes:

SolomonKurz's tweet image. Bayesian continuous-time Kaplan-Meier survival function fit with #brms and plotted in a stepped line/ribbon plot with #tidybayes:

Nana Quame repostou

Did you know that the optimal ridge penalty λ in linear regression can be *negative*? It's always strictly positive when n>p. Or when cov(x)=I. Or when true β is random. But here we argue that it can be zero or even negative when p>>n: arxiv.org/abs/1805.10939. HOW?! [1/n]

hippopedoid's tweet image. Did you know that the optimal ridge penalty λ in linear regression can be *negative*? It&apos;s always strictly positive when n&amp;gt;p. Or when cov(x)=I. Or when true β is random. But here we argue that it can be zero or even negative when p&amp;gt;&amp;gt;n: arxiv.org/abs/1805.10939. HOW?! [1/n]

Nana Quame repostou

BDA FREE (Bayesian Data Analysis now available online as pdf) statmodeling.stat.columbia.edu/2020/04/06/bda…


Nana Quame repostou

This is GAME CHANGER. Abbott to market, starting next week, a fast point-of-care #coronavirus test, delivering positive results in 5min and negative results in 13min. Will deliver 50K tests/day to start. Kudos to Abbott and FDA’s Jeff Shuren and team at CDRH who are in the fight.


Nana Quame repostou

What does it mean that 20% of hospitalizations are in patients 20-44 years? An EPI 101 tutorial on how to interpret this widely circulated figure.

nataliexdean's tweet image. What does it mean that 20% of hospitalizations are in patients 20-44 years?

An EPI 101 tutorial on how to interpret this widely circulated figure.

Nana Quame repostou

New preprint on penalized estimation of flexible #HMMs for time series of counts w/o the need to specify a (parametric) family of distributions (w/ applications to #earthquakes 📉 and #sharks 🦈): arxiv.org/pdf/1901.03275…!

_timoadam's tweet image. New preprint on penalized estimation of flexible #HMMs for time series of counts w/o the need to specify a (parametric) family of distributions (w/ applications to #earthquakes 📉 and #sharks 🦈): arxiv.org/pdf/1901.03275…!

Nana Quame repostou

The Ivory Coast and Ghana, which produce over 60% of the world’s cocoa, have banded together to charge a premium. Prepare to pay more for chocolate. on.wsj.com/2FvcjzE


Nana Quame repostou

AI curriculum 🤖 CS231n: CNNs for Visual Recognition, Stanford | Spring 2019 cs231n.stanford.edu CS224n: NLP with Deep Learning, Stanford | Winter 2019 web.stanford.edu/class/cs224n/i… CS285: Deep Reinforcement Learning, UC Berkeley | Fall 2019 youtube.com/playlist?list=…


Loading...

Something went wrong.


Something went wrong.