I am happy to announce that I made a group for all Persian Kagglists around the globe. I made this group wishing to gather all Persian Kagglist and interested people under the same roof. Communication is the key! This group would help people to find and know each other. This will also help to increase the chance of being seen for finding a desired data science job. Please spread the word. Thanks!
#kaggle #datascience
#kaggle #datascience
Linkedin
Sign Up | LinkedIn
500 million+ members | Manage your professional identity. Build and engage with your professional network. Access knowledge, insights and opportunities.
👍4
Mulitple #Postdoc (Paris, France)
Analyze complex epidemic data with transmission models
https://iddjobs.org/jobs/postdocs-to-analyze-complex-epidemic-data-with-transmission-models-2e79c7cf-1f1b-4968-aa0d-a93f8338352b
Analyze complex epidemic data with transmission models
https://iddjobs.org/jobs/postdocs-to-analyze-complex-epidemic-data-with-transmission-models-2e79c7cf-1f1b-4968-aa0d-a93f8338352b
iddjobs.org
IDDjobs — Postdocs to analyze complex epidemic data with transmission models — Institut Pasteur
Find infectious disease dynamics modelling jobs, studentships, and fellowships.
AI4Science #Call for fellows!
Deadline: Oct. 31, 2022
The #Fellowship gives researchers the chance to collaborate EPFL faculties on accelerating the use and furthering the understanding of #MachineLearning methods in the scientific discovery process. You hold a PhD, in any field relevant to this endeavour, are fluent in English and have a collaborative spirit.
https://www.epfl.ch/research/domains/cis/call-for-fellows/
Deadline: Oct. 31, 2022
The #Fellowship gives researchers the chance to collaborate EPFL faculties on accelerating the use and furthering the understanding of #MachineLearning methods in the scientific discovery process. You hold a PhD, in any field relevant to this endeavour, are fluent in English and have a collaborative spirit.
https://www.epfl.ch/research/domains/cis/call-for-fellows/
EPFL
Call for fellows
The EPFL AI4Science Fellowship gives outstanding and highly motivated young researchers the chance to collaborate with a broad range of EPFL faculty on accelerating the use and furthering the understanding of machine learning methods in the scientific discovery…
👍3
Spontaneous Symmetry Breaking
A little thread by Michael Sentef
Did you know that there was a connection to nonequilibrium physics?
Take a collection of bar magnets (it's all classical physics), with a force between neighboring ones that wants to align them with one another. We call this "ferromagnetic exchange". Now, each little magnet is free to rotate in space. So is the collective magnetization.
This is the energy of the system. x is the spatial location of the "spin", and delta is a vector connecting nearest neighbors. J is the ferromagnetic exchange interaction. The collective magnetization M is the sum of all spin vectors. As written above, the energy is minimal when all spins are aligned, and the modulus |M| is maximal then. But where will the M-vector point in space? This is *impossible* to predict from our theory. Why?
The energy functional written above does not tell us! This is precisely what we call spontaneous symmetry breaking. And this is what we all learn in our textbooks.
Now, here is a plot twist. We also learn in statistical mechanics how to compute thermal expectation values of stuff. Take all states (all possible spin configurations). Take their magnetization (=sum of little spins). Attach a "Boltzmann weight". Sum over all states. This statistical <M> value will always be zero! Why? Because the E-functional does not tell us which direction M should point into. There is no preferred direction.
In other words, stat-mech tells us that spontaneous symmetry breaking does not exist. Ummm... but you told us that it did, right? Also, how does my fridge magnet work if stat-mech forbids it to exist?
Here is an answer: We should consider the "thermodynamic limit". Which is "singular". How so? The prescription that we still all learn is the following: Pretend that there is a preferred direction, given by this little h-field (an "external magnetic field" -- where it comes from doesn't matter here).
Now all spins align with this h-field, and with one another. All problems solved? Not quite. We have to consider two limits. We can first let h->0 and then the number of little magnets N->infinity, or vice versa. The order of limits matters! (They do not "commute"). When we let h->0 for any finite system, <M> will be zero. When we first let N->infinity (the thermodynamic limit) and then h->0, <M> will remain finite and maximal, and will remember the direction in which h pointed originally. This is still all well known.
But here is a conundrum. We think of spontaneous symmetry breaking only in the thermodynamic limit. It is being taught like this everywhere. But ... nothing is infinite in nature! Yet we have fridge magnets, and they work! Why? Nonequilibrium comes to the rescue! How quickly does a fridge magnet de-magnetize after it has been magnetized? I.e., how does the limiting process in which we apply the h-field for a finite (but large) collection of spins, then turn the h-field off, work in real time? When the number of spins is large, they need to "fluctuate" away from the perfectly aligned configuration spontaneously in order to de-magnetize. This is increasingly unlikely when (a) the temperature is low and (b) N is large.
1/2
A little thread by Michael Sentef
Did you know that there was a connection to nonequilibrium physics?
Take a collection of bar magnets (it's all classical physics), with a force between neighboring ones that wants to align them with one another. We call this "ferromagnetic exchange". Now, each little magnet is free to rotate in space. So is the collective magnetization.
This is the energy of the system. x is the spatial location of the "spin", and delta is a vector connecting nearest neighbors. J is the ferromagnetic exchange interaction. The collective magnetization M is the sum of all spin vectors. As written above, the energy is minimal when all spins are aligned, and the modulus |M| is maximal then. But where will the M-vector point in space? This is *impossible* to predict from our theory. Why?
The energy functional written above does not tell us! This is precisely what we call spontaneous symmetry breaking. And this is what we all learn in our textbooks.
Now, here is a plot twist. We also learn in statistical mechanics how to compute thermal expectation values of stuff. Take all states (all possible spin configurations). Take their magnetization (=sum of little spins). Attach a "Boltzmann weight". Sum over all states. This statistical <M> value will always be zero! Why? Because the E-functional does not tell us which direction M should point into. There is no preferred direction.
In other words, stat-mech tells us that spontaneous symmetry breaking does not exist. Ummm... but you told us that it did, right? Also, how does my fridge magnet work if stat-mech forbids it to exist?
Here is an answer: We should consider the "thermodynamic limit". Which is "singular". How so? The prescription that we still all learn is the following: Pretend that there is a preferred direction, given by this little h-field (an "external magnetic field" -- where it comes from doesn't matter here).
Now all spins align with this h-field, and with one another. All problems solved? Not quite. We have to consider two limits. We can first let h->0 and then the number of little magnets N->infinity, or vice versa. The order of limits matters! (They do not "commute"). When we let h->0 for any finite system, <M> will be zero. When we first let N->infinity (the thermodynamic limit) and then h->0, <M> will remain finite and maximal, and will remember the direction in which h pointed originally. This is still all well known.
But here is a conundrum. We think of spontaneous symmetry breaking only in the thermodynamic limit. It is being taught like this everywhere. But ... nothing is infinite in nature! Yet we have fridge magnets, and they work! Why? Nonequilibrium comes to the rescue! How quickly does a fridge magnet de-magnetize after it has been magnetized? I.e., how does the limiting process in which we apply the h-field for a finite (but large) collection of spins, then turn the h-field off, work in real time? When the number of spins is large, they need to "fluctuate" away from the perfectly aligned configuration spontaneously in order to de-magnetize. This is increasingly unlikely when (a) the temperature is low and (b) N is large.
1/2
The system does not explore its state space. More precisely, the state space is not explored according to the "Boltzmann weights" mentioned above. We call this behavior "non-ergodic". Relaxation to a non-magnetic configuration is super slow for your fridge magnet. Wait long enough, and it will fall off the fridge. But "long" means "almost till eternity", provided that the magnet is sufficiently strong and sufficiently large. This is what we mean by "thermodynamic limit" for all practical purposes. The time scale on which we see "thermalization" is really really large. Now, this was all somewhere in the back of my mind, but I never thought about it so clearly. Who did? These guys: Aron J. Beekman, Louk Rademaker, Jasper van Wezel. Props to Aron, Louk and Jasper for spelling out what one rarely finds spelled out in textbooks.
I highly recommend reading their lecture notes. I should do it as well.
https://twitter.com/sentefmi/status/1572463208262434817
2/2
I highly recommend reading their lecture notes. I should do it as well.
https://twitter.com/sentefmi/status/1572463208262434817
2/2
scipost.org
SciPost: SciPost Phys. Lect. Notes 11 (2019) - An introduction to spontaneous symmetry breaking
SciPost Journals Publication Detail SciPost Phys. Lect. Notes 11 (2019) An introduction to spontaneous symmetry breaking
Watch two full days' worth of talks from our 2022 #ScienceOfScience workshop
https://youtube.com/playlist?list=PLZlVBTf7N6GqpWzuv0IjFtgwXhBsF4XI6
https://youtube.com/playlist?list=PLZlVBTf7N6GqpWzuv0IjFtgwXhBsF4XI6
Applications for our 2023 #PhD programme in Theoretical Neuroscience and Machine Learning are now open! Fully funded regardless of nationality.
ℹ️ Info and how to apply: http://ucl.ac.uk/gatsby/study-and-work/phd-programme
⏰ Deadline: 13 November 2022
ℹ️ Info and how to apply: http://ucl.ac.uk/gatsby/study-and-work/phd-programme
⏰ Deadline: 13 November 2022
A series of 14 seminars on the most influential papers of Parisi will give the opportunity to dive into the history of disordered systems (and beyond),
https://sites.google.com/gssi.it/giorgioparisiseminars
https://sites.google.com/gssi.it/giorgioparisiseminars
A series of 14 seminars on the most influential papers of Parisi will give the opportunity to dive into the history of disordered systems (and beyond),
https://sites.google.com/gssi.it/giorgioparisiseminars
https://sites.google.com/gssi.it/giorgioparisiseminars
👍1
The data science and AI division at CSE is recruiting a #PhD student in machine learning for a project on the efficient generalization using causality and auxiliary information.
https://www.chalmers.se/en/about-chalmers/Working-at-Chalmers/Vacancies/Pages/default.aspx?rmpage=job&rmjob=10849&rmlang=UK
https://www.chalmers.se/en/about-chalmers/Working-at-Chalmers/Vacancies/Pages/default.aspx?rmpage=job&rmjob=10849&rmlang=UK
Complex Systems Studies
School on Information, Noise, and Physics of Life 19-30 Sep 2022, Nis - Serbia https://indico.ictp.it/event/9826/
The lectures of the School on Information, Noise, and Physics of Life are available on the ICTP YouTube Channel at
https://www.youtube.com/playlist?list=PLRwcSE2bmyByMkUnPYXcrGFQ3-a94ghrV
https://www.youtube.com/playlist?list=PLRwcSE2bmyByMkUnPYXcrGFQ3-a94ghrV
YouTube
School on Information, Noise, and Physics of Life | (smr 3736) - YouTube
PostDoc-ERCSocsemics2022-CSS.pdf
130.4 KB
Postdoctoral fellowship (18 months, possibly extendable) in computational social science in Berlin.
Predicting friendships and other fun machine learning tasks with graphs – Feature Column
https://mathvoices.ams.org/featurecolumn/2022/10/01/predicting-friendships-and-other-fun-machine-learning-tasks-with-graphs/
https://mathvoices.ams.org/featurecolumn/2022/10/01/predicting-friendships-and-other-fun-machine-learning-tasks-with-graphs/
Feature Column
Predicting friendships and other fun machine learning tasks with graphs
Social media platforms connect users into massive graphs, with accounts as vertices and friendships as edges… Predicting friendships and other fun machine learning tasks with graphs Noah Gian…
👍2
Hello Quantum World! A rigorous but accessible first-year university course in quantum information science
https://arxiv.org/abs/2210.02868
https://arxiv.org/abs/2210.02868
👍4
Master of Science program in Social Data Science (MS SDS)
The program will provide US and Austrian degrees simultaneously to educate the future generations of data scientists sensitive to social problems.
https://networkdatascience.ceu.edu/msc-social-data-science
The program will provide US and Austrian degrees simultaneously to educate the future generations of data scientists sensitive to social problems.
https://networkdatascience.ceu.edu/msc-social-data-science
Two threads on calcalus in non-integer dimensions and how it is used in Quantum Field Theory
https://twitter.com/martinmbauer/status/1579221233593651200
https://twitter.com/martinmbauer/status/1579221233593651200
Twitter
Two threads on calcalus in non-integer dimensions and how it is used in Quantum Field Theory
Part II : Fractional dimensions in QFT
A particle in QFT is associated with a fluctuation of a field that permeates all space and time. 1/14
Part II : Fractional dimensions in QFT
A particle in QFT is associated with a fluctuation of a field that permeates all space and time. 1/14
Two open #PhD positions in Sustainability and Complex Systems with us. They may be assigned to any of seven predefined but very broad research areas including transport, energy system modeling, sustainable transitions, sustainable consumption, land use, etc
https://www.chalmers.se/en/about-chalmers/Working-at-Chalmers/Vacancies/Pages/default.aspx?rmpage=job&rmjob=10996&rmlang=UK
https://www.chalmers.se/en/about-chalmers/Working-at-Chalmers/Vacancies/Pages/default.aspx?rmpage=job&rmjob=10996&rmlang=UK
The international #internship program, CaCTüS, offers paid research internships at the #Max_Planck_Institute for Biological Cybernetics, the Max Planck Institute for Intelligent Systems and the AI Center in Tübingen & Stuttgart (Germany) to students who face significant constraints in their pursuit of a career in AI or brain research.
About the CaCTüS Internship program
cactus-internship.tuebingen.mpg.de
The sciences of biological and artificial intelligence are rapidly growing research fields that need enthusiastic minds with a keen interest in solving challenging questions. The Max Planck Institutes for Biological Cybernetics and Intelligent Systems as well as the AI Center in Tübingen & Stuttgart (Germany) offer up to 10 students at the Bachelor or Master level paid three-months internships during the summer of 2023. Successful applicants will work with top-level scientists on research projects spanning machine learning, electrical engineering, theoretical neuroscience, behavioral experiments and data analysis. The CaCTüS Internship is aimed at young scientists who are held back by personal, financial, regional or societal constraints to help them develop their research careers and gain access to first-class education. The program is designed to foster inclusion, diversity, equity and access to excellent scientific facilities. We specifically encourage applications from students living in low- and middle-income countries which are currently underrepresented in the Max Planck Society research community.
Application deadline: 4 December 2022
About the CaCTüS Internship program
cactus-internship.tuebingen.mpg.de
The sciences of biological and artificial intelligence are rapidly growing research fields that need enthusiastic minds with a keen interest in solving challenging questions. The Max Planck Institutes for Biological Cybernetics and Intelligent Systems as well as the AI Center in Tübingen & Stuttgart (Germany) offer up to 10 students at the Bachelor or Master level paid three-months internships during the summer of 2023. Successful applicants will work with top-level scientists on research projects spanning machine learning, electrical engineering, theoretical neuroscience, behavioral experiments and data analysis. The CaCTüS Internship is aimed at young scientists who are held back by personal, financial, regional or societal constraints to help them develop their research careers and gain access to first-class education. The program is designed to foster inclusion, diversity, equity and access to excellent scientific facilities. We specifically encourage applications from students living in low- and middle-income countries which are currently underrepresented in the Max Planck Society research community.
Application deadline: 4 December 2022
👍9
"Martingales for Physicists" (by Édgar Roldán, Izaak Neri, Raphael Chetrite, Shamik Gupta, Simone Pigolotti, Frank Jülicher, Ken Sekimoto): arxiv.org/abs/2210.09983
"We review the theory of martingales as applied to stochastic thermodynamics and stochastic processes in physics more generally."
"We review the theory of martingales as applied to stochastic thermodynamics and stochastic processes in physics more generally."
#PhD and #Postdoc Researcher positions (f/m/d) | Nonequilibrium statistical physics of active matter and living systems - GÖTTINGEN
https://www.mpg.de/19382858/postdoctoral-researcher-positions1
https://www.mpg.de/19382858/postdoctoral-researcher-positions1
Max-Planck-Gesellschaft
Postdoctoral Researcher positions (f/m/d) | Nonequilibrium statistical physics of active matter and living systems
In the Department of Living Matter Physics we seek to fill a number of Postdoctoral Researcher positions in the areas of non-equilibrium statistical physics of active matter and living systems