๐New paper featured in @NatureComms !
Many THX again to @MConstanceCorsi and @F_DeVicoFallani for their trust and commitment in the project.
Would you take a hint of darkside in your multilayer networks ?
๐
https://t.co/mcGKTKDQ7S
I now hold a #phd from @Sorbonne_Univ_ !
Thanks to all my colleagues at the @AramisLabParis for this wonderful PhD journey.
Link to the thesis manuscript: https://t.co/hcuxwnfzMo
Stay in touch for the next steps of my academic journey ๐!
1 month ago, I defended my #phd entitled "Characterization of multilayer networks: theory and applications to the brain", under the memorable supervision of @F_DeVicoFallani. Thanks to @_AlexArenas@LindaDouw, @alainbarrat, @gin_bianconi for our insightful discussions.
Let us explore together the dark side of multilayer networks at 3.45 pm (HS 1 - Network structure) under the sun of Vienna - contributed talk !โ๏ธ @netsci2023@F_DeVicoFallani#talks#multilayer
Which description best discriminate MEG activity of Alzheimer's disease patients as compared with the healthy controls one ? ๐ง
The dark side seems promising in this way.
How about we explore it further ?๐ง
ArXiv submission !๐
Get ready to explore the dark side of multilayer networks !
Thx so much to @MConstanceCorsi , @F_DeVicoFallani for their invaluable contributions at different stages of this work !
๐งต
https://t.co/VD4EPABbuY
We exploit the complementarity of the two descriptions to provide a local connectivity-based characterization of very diverse multiplex networks.
Actually, we need both descriptions to properly classify groups of multiplex networks !
One week ago, the in-person @APSMeetings 2023 ended. Warm thx to all of the organizers and volunteers that made it possible. Special thx to @GuidoCaldarelli for chairing the session where I presented. Lots of new ideas to explore, now !
#APSMarch2023#Physics#Science#Networks
Make sure to be present in Room 125 @APSMeetings for a Focus session on Network Theory and Applications. My talk on node-layer duality in multilayer networks is expected at 9:36 a.m PT.
See you there ! #apsmarch#physics#network
DAY 24 - "The structural neural substrate of subjective happiness", Wataru Sato et al., Sci. Rep. (2015)
https://t.co/FhkPNftNqo
Last candy from the advent article event. Where is located the subjective happiness?
Thx for following this 1 month journey !
Merry Xmas !๐
Cheers !
For the scientific, technical reading aficionados, why not reading the following article by Ju-Hyun Lee et al., Nat. Neurosci. (2022) ?
https://t.co/8NCaQfzntR
/end
DAY 23 - "What Causes Alzheimer's ? Scientists are Rethinking the Answer.", @yasemin_sap , @QuantaMagazine (2022)
https://t.co/9PCKCtWEIE
A magazine article today. Learnt a lot on current Alzheimer's theories. Nice storytelling !
Enjoy, Xmas eve is coming !
๐งต
, train graduate student to replicate studies, update online publications to mention if results were replicated.
- Anyway moving beyond p values is simply about a paradigm shift in science activities (way beyond p values)
- Will it (bayesian) update your science practices ?
/end
DAY 22 - "Moving to a World beyond "p<0.05"", Ronald L. Wasserstein, Allen L. Schism and Nicole A. Lazar, The American Statistician (2019)
https://t.co/oAAT2GsxkH
Editorial introducing a 43 papers issue to constructively replace/reform the use of p-values in science.
๐งต
- Reforming p-values cannot be done without revolutionizing research, publishing, funding and training practices: train editors and funding committees on new statistical practices, generalized "result-blind" pre-registered method ...
- No blind applications of network metrics especially in deriving null models and again, assess uncertainties on them
- (Aside) Validate method on synthetic data.
- Bayesian inference can endorsed the role of a principled based "common language" of network scientists.
/end
DAY 21 - "Statistical inference links data and theory in network science", Leto Peel, Tiago P. Peixoto and Manlio De Domenico, Nat. Commun. (2022)
https://t.co/5DkJCSpycC
Bayes law must bring network science to the next level: a milestone perspective.
๐งต
#network#inference
- Unlike many quantitative domains, uncertainties are almost never assessed, which can be highly detrimental to further analysis (Fig 1 in the paper)
- Hypothesis on the representation of networks should be explicit (with a generative model๐), ex in correlation networks...