marbletimes

6 ms·

1.
by marbletimes 5y ago
They are not slightly different, they measure something (totally) different. Confidence intervals are relative to E[y|x], Predictive intervals are relative to y. Sometimes, for example there is not much variation in y|x, the two intervals m
2.
by marbletimes 5y ago
I agree that statisticians would better than CS people appreciate the importance of uncertainty intervals--it is mostly cultural--but that "In reality there are very few ML applications that don't need confidence estimation and es
3.
The Value of Inspiration(simonevincenzi.substack.com)
2 points by marbletimes 5y ago|0 comments
4.
by marbletimes 5y ago
From a good SO answer, see https://stats.stackexchange.com/questions/16493/difference-b... "A confidence interval gives a range for E[y∣x], as you say. A prediction interval gives a range for y itself.".
5.
by marbletimes 5y ago
Correct, quantile regression is an option. Another is "pure" bootstrapping (you can see by googling something like uncertainty + machine learning + bootstrapping that this is a very active area of current research). The major prob
6.
by marbletimes 5y ago
When I was in academia, I used to fit highly sophisticated models (think many-parameters, multi-level non-linear mixed effect models) who were giving not only point estimate but also confidence and predictive intervals ("please explain
7.
by marbletimes 5y ago
But if you know that the current company cannot pay like the other company you are interviewing for, or that in this new company the future salary outlook is much better, why bother asking if they can pay you more? In the work life, one nee
8.
by marbletimes 5y ago
The assumption here is that the employee leaving the company was due to some problems that could not be solved after careful discussion with the manager, skip-level etc. But leaving a company can also happen because another company offers y
9.
Amateurs and Professionals(simonevincenzi.substack.com)
3 points by marbletimes 5y ago|0 comments
10.
by marbletimes 5y ago
Jiu-Jitsu 3 classes a week for 1.5 hours each. Weight training, 3 times a week, 1 hour each time. Running, 2 times a week, 45 minutes each run. That's on average, total 9 hours a week. Sometimes I do more, when on vacation I don't
11.
by marbletimes 5y ago
"My theory is that he’s too valuable where he is, and that they don’t want to move him elsewhere." Assuming the "they are too valuable here to be promoted elsewhere" hypothesis is true, and it may be true in some cases (
12.
by marbletimes 5y ago
The hypothesis here is that there is a trade-off between making bank and living a pleasant life, whatever attributes one wants to associate with "pleasant", maybe low professional stress, maybe time spent with loved ones, maybe wa
13.
by marbletimes 5y ago
Agreed. That's why I wrote,"but I consider that money far from ----financially---- making someone "a very successful 40-year-old"."
14.
by marbletimes 5y ago
As someone living in the United States and earning much more than I would in my home country, even though my professional contribution would be the same if the same opportunities existed in my home country (I took my PhD in my home country)
15.
by marbletimes 5y ago
It depends on where. In a European country, think Spain, Italy, France, "get good grades, study a valuable major, don't eat garbage, don't marry crazy people/idiots, stay off dangerous drugs, and mainly show up to work,
16.
by marbletimes 5y ago
What makes the real difference is curating followers and avoiding mindless "reel type" scrolling. As always, it's being intentional rather than setting hard limits (e.g., being friends only in real life, limiting the number o
17.
by marbletimes 5y ago
Feynman certainly cared a great deal about what others thought of him, and he has been known to embellish many stories to appear as the common genius in the situation. In short, he was a genius, but he also wanted it to be known very well b
18.
by marbletimes 5y ago
Let's not forget the lies that are broadcasted left and right on social media, in which thousands become millions, and trite tropes are sold like congress-worthy insights. You are right, it takes a certain type of people to succeed on
19.
by marbletimes 5y ago
It is a tricky subject. When I referred to a model I had published elsewhere, I was asked to write about the model again in the current paper, since readers may not be able to access or want to access the other paper. But more recently, I w
20.
by marbletimes 5y ago
When I was in academia--before I moved into technology and began applying statistical modeling and machine learning to a variety of tech problems--my research in mathematical biology was primarily on understanding what happens to population
21.
Power and Powerlessness(simonevincenzi.substack.com)
1 points by marbletimes 5y ago|0 comments
22.
by marbletimes 5y ago
This is one of those witty sentences that sound good (the balance of life, laughs here, tears there, if you want to enjoy living you need to accept dying, everything happens for a reason), but they are just biblical nonsense. There are ple
23.
by marbletimes 5y ago
It is one of those common tropes, "interviewing is hard", that seems to justify the inevitability of poor attitude, wasting time of interviewees (remember the interviewers are paid for those hours) and a painful process for basica
24.
The world is more accessible than you think(simonevincenzi.substack.com)
3 points by marbletimes 5y ago|0 comments