Stupid LLM Tricks, Statistics Version
Nina B. Zumel | Data Science
by Nina B. Zumel
3w ago
I recently posted elsewhere about something silly ChatGPT did when we asked it about states' names. Admittedly, that particular question was merely on the level of cocktail party conversation, or a bar bet; but it does point out that you really should never trust anything it says, unless you already know the answer. Here's what happened when we asked it a statistics question: User What are some good examples of continuous probability distribution functions with negative skew? A distribution with negative skew looks like below: tail to the left, the bulk of the distribution to the right (gene ..read more
Visit website
Clarity, Not Magic
Nina B. Zumel | Data Science
by Nina B. Zumel
1M ago
After migrating this blog to Eleventy, I made the comment on Micro.blog that Eleventy "is so much easier to wrap my arms around" than Jekyll was. That got me to wondering--why is that? After some thought, I came up with a few reasons: Newbie-friendly design and documentation that makes the process less opaque. Starter code that is simple to read, but that still illustrates useful functionality A blog deployment process that is manual, but not mystical. That last point is the crux, for me. After all, the reason I migrated my site was because the old one broke, and I didn't know how to fix it ..read more
Visit website
Blog Migration
Nina B. Zumel | Data Science
by Nina B. Zumel
1M ago
My jekyll-generated website (served by Github Pages) failed to build a couple of weeks ago; I don't know why. I'm not enough of a front-end developer to diagnose the problem. After looking around for an updated template to move to, I eventually decided to give the Eleventy static site generator a try. So this is my new site, based on Eleventy's official starter blog template. It wasn't too hard to figure out, and deployed surprisingly easily! Best of all, Eleventy is built to be stable; which hopefully means I won't run into that my website won't build, and I don't know why! problem again[1 ..read more
Visit website
Fun With Chat GPT
Nina B. Zumel | Data Science
by Nina B. Zumel
1M ago
So, my husband typed this into ChatGPT 3.5: What are some English words that are 4 letters long, start with f, and end with k? According to ChatGPT, there was only one word that fit that criterion: fork. Upon prodding, it admitted that there were maybe a few more: Typing a similar query into DuckDuckGo (4 letter word starting with f ending with k) produced a page full of links to various crossword puzzle and Scrabble dictionaries. The first link gave me seven words. A little further down, I found a link that gave me thirteen words: faik, fank, feck, filk, fink, firk, fisk, flak, folk, fork ..read more
Visit website
The What, Why, and How of AB Testing
Nina B. Zumel | Data Science
by Nina B. Zumel
1M ago
This article is a shortened version of a post from the Wallaroo Blog, originally written by Julio Barros and me. I'm posting the non-Wallaroo section of that article here, with permission, because I think it's a useful reference for A/B testing---one that I refer to myself. Hopefully, others find it helpful as well. How to A/B Test Machine Learning Models # A/B tests are a key tool of business decision-making. In this article, we'll discuss the what and why of A/B testing, and how an A/B test is designed. What is an A/B Test? # An A/B test, also called a controlled experiment or a randomized c ..read more
Visit website
New Article on the XI coefficient
Nina B. Zumel | Data Science
by Nina B. Zumel
1M ago
I have a new article up on the Win-Vector blog, about the ξ ('XI') correlation coeffient that was recently introduced by Professor Sourav Chatterjee in his paper, “A New Coefficient of Correlation”. Unlike traditional correlation coefficients, ξ does not assume that the relationship between x and y is linear; in principle, it can be any functional relationship. While this is potentially useful, ξ also has some disadvantages, too. In my Win-Vector post, I run some informal experiments to try to get a sense of what different values of ξ might mean. Please check it out ..read more
Visit website
What's Wrong with a Low(er)-Stress Job?
Nina B. Zumel | Data Science
by Nina B. Zumel
1M ago
So there's this article that's been making the rounds called "The 10 Least Stressful Jobs of 2013"; perhaps you've read it. I don't normally bother with articles like that, but it came to my attention because some of my old graduate-school friends (who are professors) threw a mini-rant on social media over the fact that University Professor is the Number One least stressful job of the year, according to the article. And just now, I tripped over a blog post where a librarian takes umbrage over the fact that they also on the list. Curiosity led me to finally skim the article in question. I won't ..read more
Visit website
On Balance
Nina B. Zumel | Data Science
by Nina B. Zumel
1M ago
One of my favorite cheesy movies is a gem from 1984 called The Adventures of Buckaroo Banzai Across the 8th Dimension. For those who haven't seen it, Buckaroo Banzai is a brilliant young neurosurgeon and particle physicist who spends his days conducting cutting-edge research. At night, he and his research colleagues -- all engineers and scientists and doctors -- rock New Jersey as a band called the Hong Kong Cavaliers. In between the brilliant science and the rock-star night life, the Cavaliers find time to save the world from an alien invasion led by none other than John Lithgow. It's very e ..read more
Visit website
On Being a Data Scientist
Nina B. Zumel | Data Science
by Nina B. Zumel
1M ago
When people ask me what it means to be a data scientist, I used to answer, "it means you don't have to hold my hand." By which I meant that as a data scientist (a consulting data scientist), I can handle the data collection, the data cleaning and wrangling, the analysis, and the final presentation of results (both technical and for the business audience) with a minimal amount of assistance from my clients or their people. Not no assistance, of course, but little enough that I'm not interfering too much with their day-to-day job. This used to be a key selling point, because people with all the ..read more
Visit website
On Writing Technical Articles for the Nonspecialist
Nina B. Zumel | Data Science
by Nina B. Zumel
1M ago
I came across a post from Emily Willingham the other day: "Is a PhD required for Good Science Writing?". As a science writer with a science PhD, her answer is: is it not required, and it can often be an impediment. I saw a similar sentiment echoed once by Lee Gutkind, the founder and editor of the journal Creative Nonfiction. I don't remember exactly what he wrote, but it was something to the effect that scientists are exactly the wrong people to produce literary, accessible writing about matters scientific. Photo: John Mount I don't agree with Gutkind's point, but I can see where it comes fr ..read more
Visit website

Follow Nina B. Zumel | Data Science on FeedSpot

Continue with Google
Continue with Apple
OR