August 2025 – Win Vector LLC
3 posts published by John Mount during August 2025| Win Vector LLC
We just finished a quick “risk in optimization series.” Here at Win Vector we think this is a neglected topic that can easily be adapted to improve and even save optimization projects. Our series was: Modeling Uncertainty For Better Outcomes: here we argue for adopting “expectation minus k standard deviations” […]| Win Vector LLC
Introduction I want to discuss how fragile optimization solutions to real world problems can be. And how to solve that. Small changes in modeling strategy, assumptions, data, estimates, constraints, or objective can lead to unstable and degenerate solutions. To warm up let’s discuss one of the most famous optimization examples: […]| Win Vector LLC
Recently here at Win Vector LLC we have been delivering good client outcomes using the Stan MCMC sampler. It has allowed us to infer deep business factors, instead of being limited surface KPIs (ke…| Win Vector LLC
This not is just a mention of a puzzle. I’ve been enjoying the Ludwig TV series. And in episode 4 they introduce a puzzle of “retrograde analysis” or “reverse chess.” …| Win Vector LLC
I have a new note on toying with the lambda calculus in Python to share here. Please check it out!| Win Vector LLC
Nina found the official solution to Dudeney’s “Digital Difficulties” (her article on it is shared here and here). The guess on how they solved it was confirmed in the Strand solut…| Win Vector LLC
At a quick glance: 32 is greater than 10. 31/32 is about 0.96875, not near pi ~ 3.141593. 31/10 = 3.1 is a worse approximation of pi than 22/7 ~ 3.142857.| Win Vector LLC
One of the bigger risks of iterative statistical or machine learning fitting procedures is over-fit or the dreaded data leak. Over-fit is when: a model performs better on training data than on future data. Some degree of over-fit is expected. A data leak is when: the model learns things about […]| Win Vector LLC
I have a new “crazy theorists” article up: “Working Through A Trivial Algorithm Whose Analysis Isn’t.” It is my notes on reading through Jonassen and Knuth’s ama…| Win Vector LLC
I have a new video demonstrating the Kelly Can’t Fail betting strategy. The idea is: this is a classroom appropriate tool for discussing allocating assets in the presence of risk. Unfortuntel…| Win Vector LLC
Introduction You may have heard of the Kelly bet allocation strategy. It is a system for correctly exploiting information or bias in a gambling situation. It is also known as a maximally aggressive…| Win Vector LLC
Introduction Data science is the study of how to work with data. It does in fact have some concepts, tools, and results. A key concept in data science project management is “the 3 Vs”: …| Win Vector LLC
Introduction Time series methods serve applications requiring inference and identification of system parameters and forecasting of future values. Box, Jenkins, Reinsel, Time Series Analysis, 4th Ed…| Win Vector LLC
Introduction To borrow a term from Hyndman, the following is going to come out as a bit of a muddle. However, that may be the honest way to survey a muddled situation. I am going to write on time s…| Win Vector LLC
I feel that the science popularizers who wrongly pushed the “1 + 2 + 3 + 4 + … = -1/12” arcania fully earned the recent “1 * 1 = 2” nonsense. To my mind the issue is: …| Win Vector LLC
Introduction I’ve been worrying a bit on how to teach thinking in the presence of uncertainty, in particular for the case of predictive models. My issue is: a lot of the “what to look f…| Win Vector LLC
I would like to write a small review of the book Bernoulli’s Fallacy by Aubrey Clayton. First the conclusion: this is a well researched and important book. My rating is a strong buy, and Bern…| Win Vector LLC
