Stanford, Harvard data science no more • Hacker News Board

Re: Stanford, Harvard data science no more

#11

Post by 2devnull » Wed, Apr 26, 2023, 2:45 PM UTC

kenjackson wrote: post #5

While I’ve led a data science team, I’ve never taken a data science course — so I’m not sure what it teaches. But i do feel pretty confident in saying that I think math does lose its usefulness around after trig. Not to say there aren’t useful aspects, but the curriculum is so inefficient. And maybe it’s because everyone needs some part of it, but that part is different for each person. Math is interesting in that th…

“math does lose its usefulness around after trig”

This is the most jarring thing I’ve read today. I can’t say I agree, but I haven’t spent 15 years studying math myself, so who am I to disagree.

Re: Stanford, Harvard data science no more

#12

Post by danielmarkbruce » Wed, Apr 26, 2023, 3:09 PM UTC

kenjackson wrote: post #5

While I’ve led a data science team, I’ve never taken a data science course — so I’m not sure what it teaches. But i do feel pretty confident in saying that I think math does lose its usefulness around after trig. Not to say there aren’t useful aspects, but the curriculum is so inefficient. And maybe it’s because everyone needs some part of it, but that part is different for each person. Math is interesting in that th…

ChatGPT might argue linear algebra and calculus are useful.

Re: Stanford, Harvard data science no more

#13

Post by CobaltFire » Wed, Apr 26, 2023, 3:18 PM UTC

mlyle wrote: post #10

CobaltFire wrote: post #6

Note that this is the recommendations for applying for Undergrad, not their classes or program offerings. Those two institutions are recommending more foundational (calculus) rather than applied courses (data science).

They continue to recommend statistics. They removed data science from a sentence that also included calculus and statistics.

Thanks, reread that. I’ve corrected it but wanted to post to make sure your comment kept context.

Re: Stanford, Harvard data science no more

#14

Post by fnordpiglet » Wed, Apr 26, 2023, 3:18 PM UTC

dooglius wrote: post #7

Geometry isn't really any more "fundamental" than statistics (for concreteness, let's say the geometry covered by the SAT and statistics as covered by the AP Statistics exam). Maybe they are using it as a proxy for formal proofs? A proof-based course in probability would actually be a lot more fundamental than either geometry or statistics, I think.

I think geometry and trig are pretty related, and have a lot of relevance in calculus especially multivariate. To your point geometry is also the first place formal proofs take shape. That said I think geometry and trig could each be a quarter of a year long and be taught to the extent needed for almost any pursuit, with supplemental at point of need. In my education geometry and trig were two entire years, and it was so dull I lost all interest in math until I took calculus. I agree a probability course would be useful, but I don’t think probability before calculus is an awesome idea. Statistics and probability can be taught at a rudimentary level without calculus, but insight really requires calculus and linear algebra. I found taking a non calc stats and probability course made the calc version harder.

Re: Stanford, Harvard data science no more

#15

Post by mrbungie » Wed, Apr 26, 2023, 3:20 PM UTC

kenjackson wrote: post #5

While I’ve led a data science team, I’ve never taken a data science course — so I’m not sure what it teaches. But i do feel pretty confident in saying that I think math does lose its usefulness around after trig. Not to say there aren’t useful aspects, but the curriculum is so inefficient. And maybe it’s because everyone needs some part of it, but that part is different for each person. Math is interesting in that th…

Wouldn't say math loses usefulness after trig. But rather, due to how it is (usually, my experience being in Chile) teached it rapidly becomes too abstract, decouples from its "real world" use cases and then it is easy to forget the forest for the trees.

The Mathematics for Machine Learning book[1] exposes this as a top-down vs bottom-up problem. While both approaches have pros and cons, a sweet spot may lay somewhere in the middle and that needs you to embrace some inevitable backtracking (i.e. college curricula should not forget to add some courses where world modelling using the math and throughfully explaining why that underlying theory and math is actually useful in describing and/or predicting reality).

PS: I also think there is still a lot of focus in resolving problems manually.

[1] https://mml-book.github.io/book/mml-book.pdf, page 13.

Re: Stanford, Harvard data science no more

#16

Post by westurner » Wed, Apr 26, 2023, 3:27 PM UTC

Do Stanford or Harvard have a UC BIDS: UC Berkeley Institute of Data Science?

> This is the [open] textbook for the Foundations of Data Science class at UC Berkeley: "Computational and Inferential Thinking: The Foundations of Data Science" http://inferentialthinking.com/ (JupyterBook w/ notebooks and MyST Markdown)

https://data.berkeley.edu/ :

> [#1 Undergrad Data Science program, #2 ranked Graduate Statistics program]

Re: Stanford, Harvard data science no more

#17

Post by vaidhy » Wed, Apr 26, 2023, 3:29 PM UTC

fnordpiglet wrote: post #14

dooglius wrote: post #7

Geometry isn't really any more "fundamental" than statistics (for concreteness, let's say the geometry covered by the SAT and statistics as covered by the AP Statistics exam). Maybe they are using it as a proxy for formal proofs? A proof-based course in probability would actually be a lot more fundamental than either geometry or statistics, I think.

I think geometry and trig are pretty related, and have a lot of relevance in calculus especially multivariate. To your point geometry is also the first place formal proofs take shape. That said I think geometry and trig could each be a quarter of a year long and be taught to the extent needed for almost any pursuit, with supplemental at point of need. In my education geometry and trig were two entire years, and it wa…

While trig is used for starting on calculus now, in practice, I never had to use trig for ML. Most of the work was in numerical differentiation and integrations. I would think trig usefulness is more in some hard sciences while ML has a much more horizontal applicability.

It is possible to teach calculus without trig (just for polynomials) and I think it is very useful just at that level.

Re: Stanford, Harvard data science no more

#18

Post by theGnuMe » Wed, Apr 26, 2023, 3:34 PM UTC

Guybrush_T wrote: post #2

I think it makes sense for them to emphasize a strong understanding of the fundamentals. It will help those students who later want to go into data science as well.

It's rote learning though as ChatGPT proves by its scores on AP calculus exams. This is just a way to maintain elite admissions.

Re: Stanford, Harvard data science no more

#19

Post by hw-guy » Wed, Apr 26, 2023, 3:34 PM UTC

kenjackson wrote: post #5

While I’ve led a data science team, I’ve never taken a data science course — so I’m not sure what it teaches. But i do feel pretty confident in saying that I think math does lose its usefulness around after trig. Not to say there aren’t useful aspects, but the curriculum is so inefficient. And maybe it’s because everyone needs some part of it, but that part is different for each person. Math is interesting in that th…

"math does lose its usefulness around after trig"

Not if you want to leave open the possibility of majoring in engineering, physical and biological science, or economics.

Re: Stanford, Harvard data science no more

#20

Post by hw-guy » Wed, Apr 26, 2023, 3:35 PM UTC

It's amazing how many bad ideas, if you scroll down far enough, are justified by an appeal to "equity." Which usually translates into dumbing things down.