tiago, to networkscience
@tiago@social.skewed.de avatar

New on the arxiv: “Network reconstruction via the minimum description length principle"

https://arxiv.org/abs/2405.010151

Short explainer thread: 1/N

@networkscience

tiago,
@tiago@social.skewed.de avatar

We need to reconstruct networks when the edges cannot be obtained directly, only their outcomes via a time series or the static states of the nodes.

Think of epidemic contact tracing, fMRI brain scans, or species abundances in an ecological system.

This is still an open problem, with many heuristic approaches out there, but few principled ones that work well.

As we had argued before, this problem is better framed as one of statistical inference of generative models:

https://nature.com/articles/s41467-022-34267-9

2/N

tiago,
@tiago@social.skewed.de avatar

We showcase this with the reconstruction of two large microbial interaction networks, from the human and earth microbiome projects. These involve more than 10⁵ species!

Have you seen reconstructed networks of this magnitude before? I haven't.

6/N

KhouryVis, to math
@KhouryVis@vis.social avatar

Congratulations to @sara_picorana for winning the @EuroVis PhD award for her dissertation "Layered graphs and their layouts, evaluations, and applications" 🎉 H/t advisor @codydunne
https://repository.library.northeastern.edu/files/neu:4f196k78t

bkeegan, to random
@bkeegan@hci.social avatar
manlius, to machinelearning Italian
@manlius@mathstodon.xyz avatar

If you had the feeling that the online discussion about COVID-19 vaccines was biased depending on the actors, you are right.

Using #MachineLearning and #NetworkScience we have shown that being a human or a bot, verified or unverified (according to previous Twitter rules) and political leaning were relevant factors for choosing the words in posts and, accordingly, the corresponding emotions to trigger.

A genuine computational social science study, led by Anna Bertani for her Msc thesis, now published also in collaboration with Riccardo Gallotti and Pierluigi Sacco

image/png
image/png

j_bertolotti,
@j_bertolotti@mathstodon.xyz avatar

@manlius This deserves a thread with an explanation for non-specialists (i.e. me 😉 )

manlius,
@manlius@mathstodon.xyz avatar

@j_bertolotti i promise I will do one once I'll get more free (tough period).

Glad you are interested.

manlius, to random Italian
@manlius@mathstodon.xyz avatar

🚨#ComplexityThoughts issue #27: #ComplexSystems #NetworkScience
Unraveling complexity: building knowledge, one paper at a time!
Not yet subscribed? It's never too late, and it's 100% free.

https://open.substack.com/pub/manlius/p/complexity-thoughts-issue-27?r=n4iyo&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true

manlius, to Engineering Italian
@manlius@mathstodon.xyz avatar

A network that shaped our present: ~400,000 km of roads connecting thousands of cities & villages.

An emblematic physical manifestation of complex adaptive systems, allowing for goods, people and cultures to flow through continents and flourish along millennia.

albertcardona,
@albertcardona@mathstodon.xyz avatar

@manlius

From this map of Roman roads one can see where the winters are icy, snowy and awful in the Iberian peninsula ... Barely any roads.

The map also shows us how much the climate has changed: modern-day Tunisia and Northern Algeria where strong agricultural lands. Quite the contrast with today.

manlius,
@manlius@mathstodon.xyz avatar

@albertcardona Exactly! That's in the keyword 'complex adaptive': since they have built a great infrastructure that at their time was satisfying their needs.

nerdsitu, to complexsystems
@nerdsitu@datasci.social avatar
manlius, to random Italian
@manlius@mathstodon.xyz avatar

Thrilled for our latest paper on network robustness & resilience, in collaboration w/ a dream team of colleagues: 👉 https://rdcu.be/dvsId

We review 2+ decades of theory & applications within a unifying framework, to better understand complex networks' structure & function.

The warmest congrats for Oriol Artime & Marco Grassia for leading this effort, and for the whole team.

It's been a long journey, with a great editorial process.

#NetworkScience #ComplexNetworks #ComplexSystems

image/png
image/png
image/png

christina, to random
@christina@social.coop avatar

Reading this from Ecological Complexity: “The adaptive cycle: More than a metaphor” and wondering if you have favorite non-metaphorical perspectives on measuring adaptive capacity in organizations and networks?





cc
@mariafarrell @robin @Valdis @dajb @ntnsndr @RuthMalan

https://www.sciencedirect.com/science/article/pii/S1476945X1830165X

junosz,
@junosz@macaw.social avatar

@robin @christina @mariafarrell @Valdis @dajb @ntnsndr @RuthMalan Bassett's work goes into a lot of network science, so markov blankets a la Friston could easily be applied, in a formal sense, to curiosity networks / knowledge graphs. A lot of this is adjacent to the central Complex Adaptive Systems literature, so there's natural connections, but not necessarily in a "grand unifying insight" sense. "Adaptive Capacity" depends in my view on model constructs

robin,
@robin@mastodon.social avatar

@junosz @christina @mariafarrell @Valdis @dajb @ntnsndr @RuthMalan Right, I'm not necessarily looking for a grand unifying insight but more for hooks to develop a more densely connected intellectual toolbox.

manlius, to random Italian
@manlius@mathstodon.xyz avatar

Quiz: can you guess what it is (in terms of a general dynamical process or even using a specific example) just by watching how the system changes over time?

video/mp4

tfardet,
@tfardet@fediscience.org avatar

@manlius from what I see there are 4 states : the naive/resting one in grey, a very brief blue state (exposed?) followed by a transition to an orange state (infectious?), then a green state (recovered?), going back to the initial naive (susceptible?) state.

So a SEIR-like model would be my guess.

manlius,
@manlius@mathstodon.xyz avatar

@tfardet great answer. It's a SEIRS, and your explanation perfectly describes it.

verukita1, to machinelearning
manlius, to Neuroscience Italian
@manlius@mathstodon.xyz avatar

is back after the summer break!

In the issue amazing new papers, from foundations to and

Unraveling complexity: building knowledge, one paper at a time!
Not yet subscribed? It's never too late, and it's 100% free.

https://manlius.substack.com/p/complexity-thoughts-issue-14

image/png
image/png
image/png

manlius, to random Italian
@manlius@mathstodon.xyz avatar

So happy and excited about this new paper...

“More is different in real-world multilayer networks”

just appeared in Nature Physics!

📚 Exactly one decade after our first paper on the math formulation of multilayer networks, this feels like an amazing milestone! 🥳

Special focus on biological networks, from proteins to brains, and on population dynamics.

Curious about a decade-long development of multilayer network science?
👉 read this (short) review about some of the most exciting points!

https://rdcu.be/dkALp

nerdsitu, to datascience
@nerdsitu@datasci.social avatar

📣 Assoc Prof wanted!

ITU is looking for an Associate Professor in and , including ,

👉 https://nerds.itu.dk/2023/08/16/assoc-prof-wanted/
Deadline: Sep 8

anna, to ai
@anna@mathstodon.xyz avatar

Hi fellow academics! I was wondering if y'all can give me some feedback on the poster I am working on for IJCAI (https://ijcai-23.org/)? I'm playing a bit with the design and how to make it attractive.

I'm attaching two screenshots. My questions are about the big eye-catcher in the middle. If you have a minute, can you please let me know what you think?

  1. Do you get the reference?
  2. Does the effort to make the reference distract you?
  3. Do you think that the text in the big pink box has NSFW vibes?
  4. Does the text in the big pink box help to pique your interest in the poster?

Thanks!

Close-up of the big fuchsia box.

anna,
@anna@mathstodon.xyz avatar

@ubi thanks! I guess that’s kind of the point, to have something stand out. If the take home message had been shorter, more to the point, no reference, do you think that that might work better for you?

ubi,
@ubi@ecoevo.social avatar

@anna It definitely stands out. I think just keep the reference. It's a bonus to anyone that gets it and it can be a conversation starter.

  • All
  • Subscribed
  • Moderated
  • Favorites
  • megavids
  • thenastyranch
  • rosin
  • GTA5RPClips
  • osvaldo12
  • love
  • Youngstown
  • slotface
  • khanakhh
  • everett
  • kavyap
  • mdbf
  • DreamBathrooms
  • ngwrru68w68
  • provamag3
  • magazineikmin
  • InstantRegret
  • normalnudes
  • tacticalgear
  • cubers
  • ethstaker
  • modclub
  • cisconetworking
  • Durango
  • anitta
  • Leos
  • tester
  • JUstTest
  • All magazines