jmount

12 ms·

151.
What Fraction of Containerization Is Just Working Around Dynamic Linking?(win-vector.com)
2 points by jmount 2y ago|1 comments
152.
Kelly Betting with Discrete Stakes(win-vector.com)
1 points by jmount 2y ago|1 comments
153.
by jmount 2y ago
I've added a follow-up note on the case of discrete stakes here: https://win-vector.com/2024/12/21/kelly-betting-with-discret... . It is known there is a dynamic programming strategy that guarantees a re
154.
by jmount 2y ago
Mathjax used to be available more places. Jupyter used to respect spacing around h2 labels. Things like that.
155.
by jmount 2y ago
That is a nice game and writeup.
156.
by jmount 2y ago
Very good point. I did some experiments and the system is very sensitive to any sort of quantization or rounding of bets. You get the expected value about the right place, but the variance goes up quickly. So in addition to your important c
157.
by jmount 2y ago
The book claims it is optimal for a set of strategies they called "sensible." I didn't think the argument flowed as well as the zero variance part of the proof, so I didn't work it in. I think the source also hinted at a
158.
by jmount 2y ago
I love that. Gaming the seed is always a possibility in demos.
159.
by jmount 2y ago
Thank you, and sorry. The Wordpress/Markdown path seems to be getting worse over time.
160.
Kelly Can't Fail(win-vector.com)
393 points by jmount 2y ago|120 comments
161.
by jmount 2y ago
Reminds me of my father in law's experience with Toyota (I think). The car's nav system had an update available on a CD you put into the car stereo. The dealership made him order one, instead of the much more reasonable loaning hi
162.
Examining Meta-Analysis(rworks.dev)
2 points by jmount 2y ago|0 comments
163.
by jmount 2y ago
Very good scenario. One variation: some researcher or division in OpenAI performs all of the above steps to get a raise. The whole field is predicated on rewarding the appearance of ability.
164.
Calculating at Pencil and Paper Scale(win-vector.com)
6 points by jmount 2y ago|0 comments
165.
by jmount 2y ago
Peter Winkler's Mathematical Puzzles is an amazing continuation with great teaching.
166.
The Monkey and the Coconuts: An Introduction to the Extended Euclidean Algorithm(github.com)
19 points by jmount 2y ago|1 comments
167.
Dudeney's Catching the Mice Puzzle(win-vector.com)
1 points by jmount 2y ago|0 comments
168.
Dudeney's Remainder Problem(win-vector.com)
1 points by jmount 2y ago|0 comments
169.
Bachet's Four Weights Problem(ninazumel.com)
2 points by jmount 2y ago|0 comments
170.
Solving 100 Bushels Using Matrix Factorization(win-vector.com)
33 points by jmount 2y ago|17 comments
171.
The ur-money glitch fraud(ninazumel.com)
2 points by jmount 2y ago|0 comments
172.
Bay Area Use R Meetup Is Back(meetup.com)
2 points by jmount 2y ago|1 comments
173.
by jmount 2y ago
plotnine is a really good grammar of graphics at this point.
174.
by jmount 2y ago
That isn't that different than choosing to charge more for accounts with credits. Make "having credits" available to an ML model and then let the ML model "ethics wash" the decision.
175.
by jmount 2y ago
The amount of data rapidly factors-out. One cat get the effect even for millions of points in 6d. It is the complexity of the model solution which is going to be degree D-1 polynomials over a shared degree D-1 denominator that drive the eff
176.
by jmount 2y ago
I think this is a big deal. In my opinion, many money making stable AI services are going to be deliberately of limited ability on limited domains. One doesn't want one's site help bot answering political questions. So this could
177.
Machine Learning Model Homotopy(win-vector.com)
66 points by jmount 2y ago|10 comments
178.
ChatGPT Is Getting Scary Better at Math(win-vector.com)
3 points by jmount 2y ago|0 comments
179.
Please Version Data(win-vector.com)
2 points by jmount 2y ago|0 comments
180.
by jmount 2y ago
It goes in phases. Anything other than pill use to be a hard no, but now we see injections moving beyond just vaccines. I've also heard from a friend that large pharma internal R&D is "pretty much just cancer for now."

More ›