Pavlos Kolias

38 posts

Pavlos Kolias banner
Pavlos Kolias

Pavlos Kolias

@pakolias

PhD | Mathematician | Statistician @ The Math.Lab Project. Lecturer in Statistics and Research Methods, University of York Europe Campus.

Katılım Eylül 2020
80 Takip Edilen20 Takipçiler
Pavlos Kolias retweetledi
Dimitris Papailiopoulos
Dimitris Papailiopoulos@DimitrisPapail·
Ten years in academia and the best part has not been what many value most ie freedom to pursue your ideas. It’s experiencing your students grow and go on to incredible trajectories. What I’ve come to know about myself is that I value permanence, presence, and people. And for all the illusions that institutions, titles, awards etc offer, none at all come close to this: watching a human absorb, even in tiny amounts, the care and effort you’ve put into trying your best to just be there for them.
English
11
24
283
17.2K
Pavlos Kolias retweetledi
Crémieux
Crémieux@cremieuxrecueil·
Interesting: If you have people sit down and watch one-minute clips of other people, they can guess how smart those people are with a correlation of r = 0.39 on average And there's a little moderation by how smart the judge is: -1 SD judges average r = 0.36; +1 SD judges, 0.42
Crémieux tweet media
English
35
15
496
32K
Pavlos Kolias retweetledi
Christos Tzamos
Christos Tzamos@ChristosTzamos·
1/4 LLMs solve research grade math problems but struggle with basic calculations. We bridge this gap by turning them to computers. We built a computer INSIDE a transformer that can run programs for millions of steps in seconds solving even the hardest Sudokus with 100% accuracy
English
248
808
6.1K
1.8M
Pavlos Kolias retweetledi
World of Statistics
World of Statistics@stats_feed·
According to a 1985 paper, everything takes 2.71 times as long as you think it will.
English
92
313
5.7K
385.9K
Pavlos Kolias retweetledi
Fermat's Library
Fermat's Library@fermatslibrary·
Stefan Banach was born exactly 133 years ago today 🎉 Here’s a cool instance of the Banach fixed-point theorem
Fermat's Library tweet media
English
36
198
1.7K
116.8K
Pavlos Kolias retweetledi
David Bessis
David Bessis@davidbessis·
How I woke up one day and decided to solve a math conjecture from the 1970s ⤵️
David Bessis tweet media
English
33
245
2.7K
592K
Pavlos Kolias retweetledi
Data Science Fact
Data Science Fact@DataSciFact·
No amount of data can persuade you of something you assign zero prior probability to. -- Cromwell's rule
English
1
12
38
4.8K
Pavlos Kolias retweetledi
The PhD Place
The PhD Place@ThePhDPlace·
When you were doing an undergrad, you wanted to do a masters. When you did a masters, you wanted to do a PhD. Now you’re doing a PhD, you want to be a cat.
English
27
193
1.1K
92.1K
Pavlos Kolias retweetledi
Anthony Bonato
Anthony Bonato@Anthony_Bonato·
Mathematicians are brutally honest, the sequel.
Anthony Bonato tweet media
English
10
212
1.5K
144.8K
Pavlos Kolias
Pavlos Kolias@pakolias·
@rmkubinec Hey there, a linear operator on a non linear function is perfectly fine. The linearity of the expectation has nothing to do with the function of RVs that you apply - it could be either linear or nonlinear.
English
0
0
0
267
Robert Kubinec
Robert Kubinec@rmkubinec·
Hmmm, what do you all think of that plot? It's showing the "average odds ratio" which means averaging logs... which is a linear operation on a nonlinear function...
Steve Stewart-Williams@SteveStuWill

Meta-analysis of field studies on gender bias in hiring: Three key findings 1. In male-dominated/gender-balanced fields, male applicants were favoured before 2009, but since then, there’s been no consistent bias or even a weak pro-female bias.

English
4
0
13
14.6K
Pavlos Kolias
Pavlos Kolias@pakolias·
@PhDemetri Yes, if you have data that show the transitions between the states you can estimate the transition probability - rate.
English
0
0
0
45
Demetri (is over at the other place too)
Demetri (is over at the other place too)@PhDemetri·
OK throwing this into the wild. Assume I observe the following discrete dynamics. Each circle is a fully observable state. The rates and transitions are arrows. This is a markov chain, basically.
Demetri (is over at the other place too) tweet media
English
5
5
22
12.4K
Pavlos Kolias retweetledi
Alex Kontorovich
Alex Kontorovich@AlexKontorovich·
An aspiring mathematician asked me recently how I spend my time. I asked if he liked video games; he said yes. I said: imagine being paid to play the hardest games you can find, all day long, and taking breaks to teach others how to play simpler levels. Every now and then, go …
PHD Comics@PHDcomics

How profs spend their time

English
30
393
5.3K
1.4M
Pavlos Kolias
Pavlos Kolias@pakolias·
@prokraustinator @PhDemetri Instead of a single Markov chain, a better alternative would be to use a Markov system or hidden Markov system, especially if you are interested in the population structure. The system assumes that you have several members that transition between the states at each time step.
English
0
0
0
232
Matt Krause
Matt Krause@prokraustinator·
@PhDemetri Entirely possible I'm underthinking it too :-)
English
1
0
0
419
Demetri (is over at the other place too)
Demetri (is over at the other place too)@PhDemetri·
OK, I need some help. Assume I have some data from a discrete dynamical system, like the one below. Arrows indicate direction of transitions in the system. How can I estimate the transition rates? Is there a modelling technique that does this?
Demetri (is over at the other place too) tweet media
English
20
3
32
19.2K
Pavlos Kolias retweetledi
Ian Hussey
Ian Hussey@ianhussey·
I've changed my mind, Bem was right, we can feel the future. It's the only explanation for how many authors I have contacted with a data sharing request have "recently wiped their laptop and lost that data". Data wiping reliably precedes data requests: precognition is real
English
3
17
85
15.3K
Pavlos Kolias retweetledi
Dr Kareem Carr
Dr Kareem Carr@kareem_carr·
Statistician here. You should not control for all relevant variables. For instance, sexism is a relevant variable but controlling for sexism would be a terrible idea.
Matt Walsh@MattWalshBlog

The "gender wage gap" when controlled for all relevant variables is one penny. Women earn 99 cents for every 100 cents men earn. Which is to say that the wage gap effectively doesn't exist. But don't let facts get in the way, Joe. You never do.

English
26
39
745
165K
Pavlos Kolias retweetledi
Smriti Mehta
Smriti Mehta@smreeteemehta·
It took my stats professor exactly 2 mins to explain the logic behind contrast matrices and I am absolutely infuriated that this has never been mentioned in any psych stats or methods class I have ever taken. This is why we can't have nice things.
Smriti Mehta tweet media
English
1
1
13
3.2K
Pavlos Kolias retweetledi
Maarten van Smeden
Maarten van Smeden@MaartenvSmeden·
How did we come to live in a world where sums of squares are too abstract and boring to teach while models with multiple hidden layers and millions of parameters are business as usual
English
17
44
459
91.8K

Keşfet

@rmkubinec @PhDemetri @prokraustinator @elonmusk @BarackObama @taylorswift13 @cristiano @BillGates