A new issue (n° 10) of #ComplexityThoughts is out!

https://manlius.substack.com/p/complexity-thoughts-issue-10

From #NetworkScience to #AI, #CollectiveBehavior & #OriginOfLife in this issue.

_________

Not yet subscribed? It's free, check it out: https://manlius.substack.com/

Want to suggest a Thought to cover for our “From the community”? Use this link: https://forms.gle/FW4EHEmfWYyXj3TFA

#ComplexSystems #BioInspiredComputing #NetworkScience #DataScience

@networkscience @complexsystems @neuroscience

An interesting article about the memory of ant colonies (and not just of ants): https://aeon.co/ideas/an-ant-colony-has-memories-that-its-individual-members-dont-have

“The Finnish myrmecologist Rainer Rosengren showed that when the [red forest] ants emerge in the spring, an older ant goes out with a young one along the older ant’s habitual trail. The older ant dies and the younger ant adopts that trail as its own [...].”.

#ants #memory #intelligence #CollectiveBehavior

Great discussion on the #Twitter exodus, with references to the classical Granovetter threshold model.

#CollectiveBehavior #SwarmIntelligence

https://link.medium.com/PpInSGj0lwb

Great discussion on the #Twitter exodus, with references to the classical Granovetter threshold model.

#CollectiveBehavior #SwarmIntelligence
https://link.medium.com/PpInSGj0lwb

#ComplexityThoughts issue #5 is going out in a few minutes!

Don't know what is it? Click on the hashtag and check it. Or follow this URL and subscribe for free: https://manlius.substack.com/

Merry Christmas and Happy 2023!

#ComplexSystems #Emergence #ComplexNetworks #SelfOrganization #CollectiveBehavior #StatisticalPhysics #ComputationalBiology #Neuroscience #BioInspiredComputing

@networkscience @complexsystems @neuroscience

Qualcuno (che in 3 anni avrebbe potuto prendere una laurea in epidemiologia, ma preferisce chiedere i vostri soldi) continua a diffondere informazioni errate, questa volta sul caso Cina.

Vediamo perchè, rispolverando un pò di storia recente e adattandola ad oggi.

1/

#ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience

@networkscience @complexsystems @neuroscience

After 6 years using Gibbsian states for network density matrices, w/ @ArshamGhavasieh we've finally shown under which conditions they are also max-entropy network states!

We gain new insights on coalescence processes and find unexpected behaviors:

👉https://arxiv.org/abs/2212.02392

#ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience

@networkscience @complexsystems @neuroscience

This is unexplored territory for us and our formalism. Collaborations are more than welcome.

CC: @lmrocha @PessoaBrain @c4computation @ricard_sole@twitter.com @cxdig@twitter.com @net_science@twitter.com

/end

@networkscience @complexsystems #ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience @neuroscience

What is the contribution of network asymmetries to the entropy?

To answer, we considered the case where such asymmetries are small enough to be considered as a perturbation of a symmetric structure.

We find the microscopic conditions to guarantee a physical entropy.

11/

@networkscience @complexsystems #ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience @neuroscience

The paper has some bonus appendices, where we show some math to better understand statistical ensemble microstates in the case of Hermitian and non-Hermitian control operators, gaining interesting insights on entropy.

But here I want to mention another appendix.

10/

@networkscience @complexsystems #ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience @neuroscience

We find that it is not trivial to tell in advance which class of network has the highest functional diversity: deducing it directly from the structure or the dynamical rules, alone, is impossible

Our density matrix formalism allows to consider the two factors, simultaneously

9/

@networkscience @complexsystems #ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience @neuroscience

It has global maxima at different propagation scales for different networks with distinct dynamical coefficients, which is higher for the original network compared to the others in almost all cases considered.

8/

@networkscience @complexsystems #ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience @neuroscience

Applications? See here the C. elegans neural network and its response to perturbations compared against two generative models from different ensembles.

We focused on the fragility of functional diversity: an
indicator of how a system can stay operative in noisy environments

7/

@networkscience @complexsystems #ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience @neuroscience

Since we are less constrained, we can analyze what happens when coupling different network topologies with distinct (nonlinear) dynamics:

- build the corresponding partition functions
- calculate the entropy, ...

6/

@networkscience @complexsystems #ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience @neuroscience

In contrast with the statistical propagator, the diag elements of this density matrix admit a probabilistic
interpretation even for non-Hermitian control operators, allowing to exploration a broader set of problems.

Nicely, it reduces to what we know in the Hermitian case.

5/

@networkscience @complexsystems #ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience @neuroscience

Again, we view the propagation of perturbations through
a complex network as a model of information flow.

However, here we use signal processing to understand how system’s units communicate with each other.

4/

@networkscience @complexsystems #ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience @neuroscience

The statistical field theory approach makes a great job when we work with diffusive dynamics on undirected networks, but it is difficult to extend to nonlinear dynamics and more complex topologies (signed, directed, ...), which lead to non-Hermitian control operators.

3/

@networkscience @complexsystems #ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience @neuroscience

If you want to get a better idea of the content, you should be familiar with the context.

I keep a fair list of papers related to the context and you can parse it before going ahead:

2/

https://twitter.com/manlius84/status/1326933394341711874

@networkscience @complexsystems #ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience @neuroscience

With some delay, here the thread on our latest paper w/ @ArshamGhavasieh where we provide an alternative approach to build network states in terms of density matrices.

We generalize the existing approach and significantly extend it.

1/

https://arxiv.org/abs/2210.16701

@networkscience @complexsystems #ComplexSystems #ComplexNetworks #CollectiveBehavior #StatisticalPhysics #NetworkStates Of interest for: #ComputationalBiology #Neuroscience @neuroscience

“Our analysis suggests that it is possible to conceive intermittent collective strategies that take advantage of both hierarchical and democratic organizational schemes.”

#ComplexSystems #CollectiveBehavior #InformationFlow #ComplexNetworks @complexsystems @networkscience https://www.nature.com/articles/s41567-022-01769-8