175 – Disordered Systems
Rate/Vote |
Guest: Peter Sollich Host: Markus Voelter Shownoter: Stefaan Rillaert
In this episode we talk with Peter Sollich of King’s College, London about disordered systems, statistical mechanics and complexity. In particular, we discuss the difference between quenched and annealing disorder, the relation to entropy, complexity and chaos, the formalisms used to tackle such systems as well as a whole lot of examples from physics and other sciences.
Introduction
0:00:30Peter Sollich | Support Omega Tau on Patreon | Vote on old episodes on the website | Disorder Systems research group at King's College London (NETADIS - Statistical Physics Approaches to Networks Across Disciplines | CANES - Cross-disciplinary Approaches to Non-Equilibrium Systems)
Definition disordered systems
0:06:20Order and disorder in physics | Crystal | Spin glass | Ferromagnetism | Antiferromagnetism | Geometrical frustration | Magnet | Quenched disorder | Annealed disorder | Mathematical optimization | Constrained optimization | Solution space | Statistical physics | Entropy | Thermal fluctuations | Chaos theory | Butterfly effect | Ergodicity | Integrable system
Approaches to tackling disordered systems
0:41:21Closed-form expression | Numerical analysis | Simulated annealing | Metropolis–Hastings algorithm | Detailed balance | Boltzmann distribution | Partition function | Boltzmann factor | Replica trick | Belief propagation | Cavity method
Non-equilibrium systems
1:00:05Non-equilibrium thermodynamics | Glass | Metallic glass | Time reversibility | Irreversible process | Steady state | CANES programme | Climate model | Bayesian statistics | Systemic risk
Equilibrium systems
1:20:20Exploration-exploitation trade-off | Mathematical model | Sphere packing | Jammed systems | Protein interaction networks
Current work
1:39:30Peter Sollich research output | Memory effects in biochemical networks | Gaussian noise
Pingback: Die letzten und nächsten 24h, Samstag, 04.07.2015 | die Hörsuppe
Great episode!
I don’t mind the social sharing reminders. I agree it’s mostly unnecessary. But sometimes it’s not, and you may as well promote yourself. I imagine much of your audience is older than 25. So their first reaction to enjoying a product probably isn’t to share it on social media like a modern teenager might.
Indeed. It’s also at the very end. So if somebody is annoyed, they can always just switch off :-)
Interesting but hard to understand. Still not sure I ‘captured’ everything :-)
Same here Bert :-)
Even if I didn’t either get everything I found this episode very interesting, especially the multi-faceted application of these kinds of systems and their common parameters.
Right. Interesting topic, hard to understand.