Joe Baron

Postdoc in Theoretical Physics of Disordered Systems at ENS Paris

[Joe Baron @JoeBaron22]

@skdh Was this really a "popular" idea? Most physicists simply accept QM for what it is

[Joe Baron @JoeBaron22]

The biggest advantage of the method is the simplicity and generality of the expressions obtained for the eigenvalue spectra. The approach is extended to a variety of other ensembles (matrices with non-Gaussian statistics, for example).

[Joe Baron @JoeBaron22]

New preprint: https://t.co/8RGcNFbIuh

A new method for calculating the eigenvalue spectra of non-Hermitian random matrices is presented. Using a path-integral approach (with Feynman diagrams), the sparse corrections to the classic elliptic and semi-circular laws are found.

[Joe Baron @JoeBaron22]

As a consequence, we show that the eigenvalue spectra of random matrices (à la Robert May) can be used to deduce the stability of “feasible” ecological communities, but only if the emergent non-Gaussian statistics of the interactions between species are taken into account

[Joe Baron @JoeBaron22]

New in PRL: https://t.co/0J78qOApt3

We show that the full statistics of the interactions of species in a surviving Lotka-Volterra community, beyond those of a Gaussian ensemble, are required to correctly predict stability, i.e. the RMT universality principle fails in this system

[Tobias Galla (also on BlueSky) @tobiasgalla]

Fully funded (v well paid) PhD studentship to work with me at @IFISC and @gwaconstable (York) on Statistical Physics of Stochastic Populations. Deadline 25th Jan. https://t.co/7eYfLmnC4M. Get in touch (email) in next few days, I can help with application to La Caixa foundation.

[Joe Baron @JoeBaron22]

New in PRE: I find the corrections to the well-known elliptic, circular and semi-circular laws in random matrix theory due to network degree heterogeneity.

We see that NDH tends to be a destabilising influence unless edge weights are very antisymmetric.

https://t.co/oZ21rvsjOA

[Joe Baron @JoeBaron22]

@dannagifford Very exciting Danna. Very well deserved. Congratulations!

[Joe Baron @JoeBaron22]

New preprint: https://t.co/h7kXjEBPMB

Using random matrix theory, the effect of network structure on the stability of a complex network is found. I derive modified versions of the well-known elliptic, circular and semi-circular laws for non-zero network heterogeneity.

[Joe Baron @JoeBaron22]

Very happy to see this work published.

By mapping the random matrix problem onto a dynamical system, we use a path integral approach to find the outlier eigenvalues of matrices with previously neglected correlations.

In the end, we arrive at a surprisingly simple formula.

[Tobias Galla (also on BlueSky) @tobiasgalla]

"Eigenvalues of Random Matrices with Generalized Correlations: A Path Integral Approach" (with @JoeBaron22, T. J. Jewell, C. Ryder)

Now available in Physical Review Letters:
https://t.co/8p0oe2GgMD

[Tobias Galla (also on BlueSky) @tobiasgalla]

Non-Gaussian random matrices determine the stability of Lotka-Volterra communities (w @JoeBaron22, T. J. Jewell and C. Ryder) https://t.co/eFKBxFiQ3p

- Interaction matrix btw species emerging from random Lotka-Volterra eqs is non-Gaussian (even if original is Gaussian).

>>>

[Joe Baron @JoeBaron22]

A lock-down project finally come to fruition!

A model of opinion dynamics with disordered interactions. Using dynamic mean-field theory, I find the kinds of interactions that promote that consensus, polarisation and a spread of opinion within this model https://t.co/6OKF4x5K6B

[Joe Baron @JoeBaron22]

@tobiasgalla All the best Tobias and a speedy recovery!

[Tobias Galla (also on BlueSky) @tobiasgalla]

Eigenvalues of random matrices with generalised correlations: a path integral approach (w @JoeBaron22, T. J. Jewell, C. Ryder)

We map the resolvent of a random matrix onto the response function of a disordered dynamical system, ...

https://t.co/pjdpU7hh3A

[Joe Baron @JoeBaron22]

I was of course gutted to see England lose by such a slim margin last night, but after having seen the disgraceful conduct of some of the England fans, I can honestly say that I am far more disappointed by their behaviour than by the result.

[Joe Baron @JoeBaron22]

@ProbFact The resulting reduced numerator and denominator are now each chosen randomly from the set of integers. We're back where we started. Again, there is a 1/4 probability of getting o-e in the reduced fraction or e-e. One obtains the series P(d reduced is even) = 1/4 + 1/16 +... = 1/3

[Joe Baron @JoeBaron22]

@ProbFact The probability of choosing o-e is also 1/4 and we get an even denominator with certainty (no factor of 2 to cancel). However, the combination e-e (again, probability 1/4) could give an even or an odd denominator after cancellation. We cancel a factor of 2....

[IFISC (UIB-CSIC) @IFISC_mallorca]

#ifisc_seminar Annabel Davies, The University of Manchester, Statistical physics approaches to network meta-analysis (28-04-2021 12:30 PM) https://t.co/n4ZJcGPyd9