#datascience cheatsheets for you

This is called "A Gentle Introduction to the Hessian Matrix"

Hessians are somewhere between #linearalgebra #calculus and #rstats but still a core aspect of #datascience

All in all, building and deriving things like these are probably only useful when developing a unique solution. For the vast majority of cases, having a general understanding is enough.

... actually, I am pretty sure that there is a #python library for just such an occasion (I have never looked though so ymmv)

... I thought that my work deals with some high level stuff but this guy is completely over my head. I swear, it is all I can do to barely keep up.

To my seniors: I do not see how you guys do it (and consistently at that) but I gain a new level of respect for you every time it happens.

On the other hand, does that mean I will stop? Why? I just told you the stuff I talk about is not beginner-friendly. Words of wisdom from my betters and those before me, "Get gud scrub"

Oh, yeah. To be clear, I have made none of the cheatsheets I share. I do not know why, but I do not recall ever bothering to make them myself. I kept most of my work, though you might be hard-pressed to use it like a formula sheet. I.e. I am in no position to ask an author to remake their sheet.

But like, for beginners, they can only be a little better than poison. Like, the stuff I post and talk about is rarely beginner-friendly, but... those things are on another level.

Like, they enforce that first hump on the Duning-Kruger curve. If you are new, they make you think you know something when you do not even yet understand how much you do not know.

But like, I already did it, though. Others did, too. Like, we already passed the Diff. Eq courses, or the calculus, linear algebra, and the programming courses. We already spent time getting intimate with probabilities and distributions (😏). So cheatsheets, as much of a crutch as they are, can be quite handy to have around.

And, tbh I would be lying of I said it was not super convenient. Like, when I am on a project, I rarely have the time (or patience) to flip through notes for the subtleties of a loss function.

Or, like, evaluating the appropriateness of using a determinant or eigenvalues to find a solution, based on some informational context.

But like, that is what they are for, right? I mean, think back to your Diff. Eq. and discrete math days. You made a formula sheet (cheatsheet) to help with exams when you did not have the luxury of deriving the entire solution.

Like, they gave you a quick way to say, "Hey, plug the result into this equation to get the answer so you can move on to the next step."

So, like, I know I share some of the (passable) cheatsheets I come across. (Some I forget about, but others are just plain bad, okay, like, a "F - -" grade for quality)

And, like, they can be quite useful... for the ones who already took the classes or have spent time learning the information.

Okay. After that bit of hilarity yesterday, have some stuff on #linearalgebra

Not a formula sheet but still useful for developing your #datascience intuition

@diff @SamiMaatta

🤣 honestly, I was in a bit of a hurry when I saw it so I did not give it a proper check... but it is so bad that (even though I am tempted to delete it) I can not help myself from finding it hilarious

Though I suppose that also means that, technically, such a cheatsheet still does not exist (a good one anyway)

I am not sure whether to be excited or frustrated that a #mathematics cheatsheet for #datascience exists.

Tbh I never even thought to make one when I was first learning this stuff. But hey, here it is.

Edit: ... so technically, a good formula sheet for it has yet to appear. The section headings and bold text ("dot in column" ???) ...have high accuracy with low precision

At the end of the day though, I kind of feel like there is a disparity between educational requirements, job duties, and employer expectations (aka the circle jerk that is sometimes called "company culture"); some (a few?) employers do have legitimate expectations though 🤔

Anyway, I also feel that disparity is the cornerstone of the (irrational?) perspective that many have toward remote work.

The thing that intrigues me is the future impact of remote work on outsourcing and worker visas like the H1B.

Tbh, I do not see the visas having much of a negative impact but outsourcing might take a huge hit. Maybe not for things like call centers but 🤷‍♂️? #ai agent(s) can fill "high school diploma" jobs but should it?

Edit: If your rebuttal derives from your work ethic mandating being surrounded by others who are also engrossed in their work, then that is understandable.

However, it is also not the subject of discussion.

Edit: If your rebuttal derives from an alignment with your mental model, then I suppose the sun will not rise and the Earth will cease rotation when you expire.

Either way, please keep your head in the sand. Its attachment to your shoulders is irrelevant.

Edit: Before anyone starts with that "perception is reality" nonsense, here are some consequences of that being true.
• Trans people do not exist. By that token, neither do homosexuals.
• The boogeyman is real. It hides in dark places, like your shadow, waiting to steal your ■■■■ from your ■■■■
• You need to be wary of kelpies. Especially on rainy days.
• Cthulhu fhtagn.

Any of those being false negates the initial premise.

Tl;dr: remote work is fine for some duties

But like, I get it. For some people, "going to the office" is an exclusive membership club where they can find companionship and placate their egos.

Abstract: "...whether working from home (WFH) affects workers’ performance in public
sector jobs. Studying public sector initiatives allows us to establish baseline estimates on the
impact of WFH net of incentives. Exploiting novel administrative data and plausibly exogenous
variation in work location, we find that WFH increases productivity by 12%. These
productivity gains are primarily driven by reduced distractions. They are not explained by
differences in quality, shift length, or task..."

🤔 this one is a bit out of my depth, but this study reveals some interesting insights about remote work. #futureofwork

Fenizia, A., & Kirchmaier, T. (2024). Not incentivized yet efficient: Working from home in the public sector.

The paper is available on Google Scholar

I mean, you guys already know I like results supported by hard data.