Dirk L.

5.1K posts

Dirk L. banner
Dirk L.

Dirk L.

@Dirque_L

Does and teaches applied mathematics. (he/him, er/ihn)

Katılım Eylül 2016
356 Takip Edilen1.2K Takipçiler
Dirk L.
Dirk L.@Dirque_L·
@bremen79 When you read preprints, you are effectively a reviewer
English
0
0
2
414
Francesco Orabona
Francesco Orabona@bremen79·
Weekly reading group: - New viral optimization paper! - Wait, this claim is clearly false... - This one is not supported by the experiments... - This formula doesn't make any sense... - This assumption is false in practice... - This was known 10 years ago... etc. etc.
English
5
0
60
4.4K
Dirk L.
Dirk L.@Dirque_L·
@wtgowers Every sign has a story behind it 😬
English
0
0
6
347
Dirk L.
Dirk L.@Dirque_L·
@majavonwestphal Wenn man es Referat nennen würde, würde das bei uns gehen (kann Ausarbeitung und Vortrag und Auseinandersetzung mit Inhalt umfassen).
Deutsch
1
0
1
65
Maja von Westphal
Maja von Westphal@majavonwestphal·
Schöne Idee. An der Uni geht das nicht, denn das wäre eine 2. Prüfung im selben Seminar, die die Prüfungsordnung verbietet. Der Unterschied zwischen Performanz und Kompetenz ist den Prüfungsämtern seltsam fremd - da sitzen halt Verwaltung und Jura, keine Bildungswissenschaft.
Robert Pondiscio@rpondiscio

A modest proposal to the solve problem of AI-generated student essays: give each student a surprise quiz based on their own papers, administered in class, without their submitted work to refer to. Use AI to generate the customized quizzes. Fair is fair.

Deutsch
6
0
15
2.9K
Dirk L.
Dirk L.@Dirque_L·
The exponential function has the remarkable property that it is the unique function f that satisfies the differential equation f'(x) = exp(x) with initial value f(0)=1.
English
0
0
2
277
Dirk L.
Dirk L.@Dirque_L·
@Ni_Zakharov I can't remember when I last did that - in the last few years no journal insisted that I provide the paper in their format.(OTOH you wouldn't believe what shitty LaTeX people write. Moreover, LaTeX isn't all that helpful when it comes to publishing (mostly html these days))
English
1
0
6
579
Nikita Zakharov
Nikita Zakharov@Ni_Zakharov·
Why is manuscript formatting according to the journal's guidelines still a thing in 2026? People, you have my LaTeX code and my replication data - what else do you want from me!?!?!?
English
7
2
50
10.7K
Dirk L.
Dirk L.@Dirque_L·
@sonukg4india Why shouldn't it? Just look at the terms of the Taylor series: For x=1 you get an alternating series with "harmonic decay" (like 1/n basically). If you plug in some x with |x|<1 you get geometric decay, so much faster convergence.
English
0
0
0
17
SKG
SKG@sonukg4india·
After using hundreds of terms of that series, I can see that it converges really slowly. However, instead of x = 1, using x = 1/sqrt(3) in the same series for arctan(x) gives a series that converges to π much faster. Why does a change of angle affect how fast it converges to π?
English
2
0
4
692
Dirk L.
Dirk L.@Dirque_L·
@NLRG_it Write out the pdes for the real and imaginary parts and you'll see a kind of wavy behavior.
English
0
0
0
356
no longer respecting gardenias
no longer respecting gardenias@NLRG_it·
if you ignore the potential term schrodinger's equation looks like a diffusion equation. so why doesnt an electron in a vacuum just diffuse into nothingness? why doesn't it look like a wave equation?
English
6
0
25
4.9K
Gro-Tsen
Gro-Tsen@gro_tsen·
I'd say a mathematician's main three fears when proving a new result are: ‣ that there is an error in the proof (the result is actually false), ‣ that there is a much simpler proof, rendering the result trivial, ‣ that the result is already known and published (so not new).
English
20
12
313
14.6K
Doc Kien
Doc Kien@BasedScience·
im talking about ideas, not 'research progress'. they can't steal research progress because they can't steal time. few people have ideas and everyone is desperate to publish - sufficinet ingredients for theft. fortunately their incompetence can cancel out, ie they steal ideas they don't understand and misrepresent them
English
2
0
0
90
Dirk L.
Dirk L.@Dirque_L·
@predict_addict This one is 14 years old and about a totally different topic.
English
0
0
0
8
Valeriy M., PhD, MBA, CQF
Valeriy M., PhD, MBA, CQF@predict_addict·
Did you know Germany has a ridiculous number of universities? 422+ higher education institutions (109 full unis + 209 applied sciences) for ~84 million people — ~5 per million. Decentralized federalism gone wild. But here’s the catch: “free” tuition (for everyone, including internationals) funded by taxpayers has led to massive overcrowding. Lecture halls packed with hundreds, PhD students doing the teaching, strained resources, and quality suffering in many places. Taxpayers foot the bill while dropout rates linger and infrastructure groans. Is spreading thin really “accessibility,” or just unsustainable populism that dilutes excellence? Compare to more selective systems — maybe fewer, better-funded unis would serve students (and the economy) better than this bloated network.
Valeriy M., PhD, MBA, CQF@predict_addict

TU Munich is basically the Lidl of elite STEM universities. They throw you in the deep end and expect you to fend for yourself: study hard, figure it out alone, and if you fail, you’re simply gone. The big difference with proper universities abroad is that TUM piles on four times the contact hours and four times the self-study on top of that. Why does Germany feel the need to Lidl-ise everything? Airports, airlines, supermarkets… and now even higher education. Sure, it looks great on the balance sheet in the short term. But the long-term price is brutal: Germany quietly bleeding competitiveness across the very domains that once defined it—math, automotive engineering, AI, you name it.

English
11
2
65
12.4K
Dirk L.
Dirk L.@Dirque_L·
@predict_addict That's 13 years old (and students numbers peaked a year after this).
English
0
0
0
6
Dirk L.
Dirk L.@Dirque_L·
@predict_addict I have taught these classes for over 15 years (and studied the subject myself before) and it's rare, but sure, AI has to be right about this. Moreover, students numbers are going down in Germany for about five years, especially in STEM fields.
English
0
0
1
9

Keşfet

@bremen79 @wtgowers @majavonwestphal @mathandcobb @disregardfiat @Ni_Zakharov @sonukg4india @NLRG_it