Actions

Complex Systems Summer School 2014-Projects & Working Groups

From Santa Fe Institute Events Wiki

Complex Systems Summer School 2014


Contents

Dynamics of Norms Under Reinforcement Learning

Dynamics of Reinforced Norms

Summary:
We investigate how societal/cultural norms are reinforced and how they affect decision making of an agent. Agent decision is conditioned by an identity saliency vector. Each identity of the agent connects it to a number of other agents that reinforce the choice of the agent. For instance, the role of agent as a daughter connects it to her family which may support decision to invest in education or not. We study how the the connection of agent with other agent may affect the saliency vector and cause the agent to change the decision rule.

Why do we study this?
The application of this project is to inform policy making and designing the right intervention points (pressure points) that one can use to change norms in a society.


Interested:
Francesca
Alireza
Brais
Leo

Bitcoin

Bitcoin Page

Bitcoin is an online peer-to-peer currency which has gained popularity in recent years. With increased general interest and now more and more companies accepting transactions in Bitcoin, so to has its value and is currently priced at $633 for 1 Bitcoin!

So what is Bitcoin?

The idea behind Bitcoin was to remove the need for a centralised banking system. The way this works is to have all transactions made public and then have the Bitcoin community validate transactions by donating computational power (and rewarded in Bitcoins). For further details see this blog:

http://www.michaelnielsen.org/ddi/how-the-bitcoin-protocol-actually-works/

So why is it interesting?

Usually companies and individuals keep all of their financial records private, but ALL transaction in Bitcoin are publicly available, so this provides a unique opportunity to study financial transactions. While the transactions are public, the owners of Bitcoin remain (relatively) anonymous, and so this has led to associations with criminal activities (e.g. on the silk road) from illegal drugs to hired hit-men. In addition to illegal activities using the currency, there have also been a number of illegal activities against the currency, such as high valued heists of the order of $100 million.

Oh, and of course its a big temporally evolving network.

If you're interested in working on this then add your name below. If there is interest, then we can have a "Bitcoin, beers and blackboard" session to throw some ideas around.

Interested:

Leto
Shai
Flavia (flamarquitti@gmail.com)
Jessica Santana jsant@stanford.edu
Stefan
Ana María
Brais (brais.alvarez@eui.eu)
George(qiaozhi827@gmail.com)
Sean Hayes (shaye004@ucr.edu)
Matthew Ayres (matthew.ayres@growthandinnovation.com.au)


Meeting today (Jun 12) 6pm in the main conference room.

Resources: https://bitcoin.org/en/resources
http://blockchain.info/charts
http://blockexplorer.com/
http://bitcoincharts.com/
Web of trust: http://bitcoin-otc.com/trust.php

MITRE Data Sets

To access the data please contact Juniper she has it on a hard drive. Here is a PDF that explains the datasets and gives some sample challenge questions. MITRE DATA PDF If you have any specific questions about the data you can contact Matt Koehler at mkoehler@mitre.org


Fractal-like structures in economic data

In the 1960-70s Mandelbrot showed that some economic time series have fractal-like structures, i.e. they look the same at many time scales. The existence of these structures has been debated since. Do economic time series like S&P 500 index have fractal-like structures? If yes, how fractal-like are they?


Interested

Vipin (vipin.veetil at gmail dot com)
Blaz
Qiao Zhi
Leo
Cornelia (cornelia.metzig@hotmail.com)

Microbial Community Data Sets

The Earth Microbiome Project EMP is a massively multidisciplinary effort to analyze microbial communities across the globe. The general premise is to characterize the Earth by environmental parameter space into different biomes and then explore these using samples currently available from researchers across the globe. All data sets are processed in the same way (DNA extraction, PCR primers, sequencing, bioinformatics), making them inter-comparable. You can explore these data sets (including some time series, and a bunch of spatial samplings) at the following link EMP Data (no need to create a login ID, just scroll down to 'Download Public Data'). If you have questions, please contact Sean Gibbons (sgibbons at uchicago dot edu).


The pollination problem from a game theory approach

The old title was: "Does Larger Memory Capacity Brings about Evolutionary Advantage?"

Pollination systems are amazing. The expected from a flower visitor , like a bee, is that it enters the flower looking for its resource (nectar) and when gets out, it is carrying many pollen grains in its body. Then, in the next visiting flower looking for more resource, it eventually pollinates this second flower with pollen bind on the body. But what happens is that this system can face up some free-riders, interested in getting their nectar quickly and easily. Yes, they exist! The nectar robbers (or cheaters) can rob nectar in different ways. One of the most amazing techniques is making a hole in the basal part of the flower and getting the nectar without touching the reproductive parts of the flower.

However, if you think that plants always evolve to avoid this cheaters, maybe you are very mistaken. The presence of the cheaters reduces the amount of nectar inside the flower, and when a pollinator visits this flower after the cheater, it must look for more flowers to get the necessary amount of nectar. It potentially increases the outcrossing reproduction, very necessary for some plant species (and it can also produce a higher fitness than the autopollination system in many other plants which can reproduce via both systems)

We plan to model this unexpected history using a game theory approach, with some flavor of Public Goods game, incorporating memory in the path made by the plant visitors.

We also still consider our initial idea of studying the influence of more/less memory on the qualitiy of evolved strategies in an evolutionary game theoretic context. The focus would be, however, on the fact that in nature more memory comes with a higher cost.

Interested:
Degang Wu (dwuab@ust.hk)
Blaz
Brian
Flavia (flamarquitti@gmail.com)
Cole Mathis (cole.mathis@asu.edu)
Claudius
Luis Martinez (fnxabraxas@gmail.com)

North American Breeding Birds Survey and tropical trees

This dataset contains ~4500 sites where populations of birds (~600 species overall, ~60 species on average in every site) were sampled over the past 44 years. This gives numerous time series of both population sizes and the overall number of species. Some problems with this dataset include large observational errors. A dataset of tropical trees with the exact diameter, identity (from among 300 species) and location of ~250000 trees over 6 censuses is also available. Please contact Michael Kalyuzhny for these.

The Multiplex Networks

Multiplex Networks Page

How does the structure of social networks affect the emergence and persistence of norms? Why are some norms (like fashion styles) less persistent than others (like religious beliefs)? Is this because different kinds of norms live on different social networks (with same individual participating in different networks)? If yes, how do these social networks interact?

And finally, what do the answers to the above questions tell us about policy interventions? Can certain critical properties of network structures be exploited to change norms, like going to the moon without much fuel? If two networks interact, say religious belief and fashion styles, can interventions in one be used to bring about changes in another?

PS: Here link 1 is a short review of multiplexes that I put up together. It is not exhaustive but the references can give us an idea of the field. If you find any error please let me know (Massimo).


Interested

Vipin (vipin.veetil at gmail do com)
Alberto (alberto.antonioni@unil.ch)
Francesca
Nhat (nhattdnguyen@gmail.com)
Sanja (sanjakojasanja@gmail.com)
Massimo
Alireza (alireza.goudarzi at gmail.com)
Claire (lagesse.claire at gmail.com)
Cecilia (ci.andreazzi@gmail.com)
Sean Hayes (shaye004@ucr.edu)
Brais (brais.alvarez@eui.eu)
Matthew Ayres (matthew.ayres@growthandinnovation.com.au)

Network Tolerance of Failure

Network Tolerance Page

How might a network endure non-catastrophic failure without isolating the failing components? Most network failure models consider resilience against failure as a result of isolating failing components. In contrast, is it possible for the network to be robust through "tolerance" of failure? Perhaps, for example, a symbiotic relationship sustains a weakened node until it has recovered its prior performance. Or perhaps a transmitting network retains a connection to an offline node to reduce an anticipated memory load of re-establishing the connection when it comes back online. What are some of the ways in which networks are able to maintain connection to a failing node without failing themselves in the process - how is the cascade halted without isolation? This is meant to be a broad question to generate more specific ideas. Importantly, this question refers to "tolerance of" failure in contrast to "resilience against" failure.
Contact: Jessica Santana jsant@stanford.edu

=> It may be interesting to at least discuss (if not merge) this with the project we're proposing (not yet on here but it involves the optimal design of the isolation approach as a function of a measure comprising the network topology, node properties, node-saving attributes of a regulator etc.). There's also the last project on this page dealing with topology and resilience which could also be incorporated. - Nicolas

Interested:

Junjian Qi (junjian.qi.2012@ieee.org)
Ells Campbell
Leto
Alireza (alireza.goudarzi at gmail.com)
Claire (lagesse.claire at gmail.com)
Sean Hayes(shaye004@ucr.edu)
Ali Kharrazi
Nicolas K. Scholtes
Cecilia (ci.andreazzi@gmail.com)
Lin Li (linnlii2495 at gmail dot com)

  • For those interested in this project, we will be meeting at 1 pm in the cafeteria with a related project to determine how to split up the groups. Cheers, Jessica*

Growth of Cities

Cities Growth Page

Do foraging animals and growing cities utilize resources in the same way? We're interested in building an agent-based model which generates a road network on a map of varying resources by following a set of simple, probabilistic rules. How do the properties of this network evolve through time? How much of city growth can be explained by resource constraints? Do simple rules of growth parallel simple rules of animal foraging behavior? This project will explore agent-based modeling, but will also present opportunities to examine the limits of modeling. Contacts: Diana LaScala-Gruenewald (dianalg11 at gmail.com) and Claire Lagesse (lagesse.claire at gmail.com).

Interested:
Morgan Edwards (morgane@mit.edu)
Rohan Mehta (rsmehta at stanford dot edu)
Alberto Antonioni (alberto.antonioni@unil.ch)
Ernest Liu (yu.liu@math.uu.se)
Michael Kalyuzhny (michael.kalyuzhny at mail.huji.ac.il)
Hiroshi Ashikaga (hashika1@jhmi.edu)
Bernardo Furtado bernardo.furtado at ipea.gov.br
Alex Brummer brummera@email.arizoa.edu
James Holdener (jholdener@mitre.org)

Towards a Unified Theory of Biodiversity

The idea is to build on the Unified Neutral Theory of Biodiversity and create a realistic unified theory that incorporates important scaling phenomena following power laws, energetic constraints, and stochasticity that have been previously neglected. Basically, we want to modify the unrealistic assumptions that birth, death, speciation rates are stochastic uniform functions across all species within a given meta-community. In fact, these processes have recently been shown to scale with body size following power laws of the form Y= C * M^alpha, where C is a constant independent of body mass, M, and alpha is the scaling exponent. These constraints will also dictate how much energy is required at different trophic levels and body sizes, ultimately constraining abundance of organisms in natural systems. Stochasticity will still play a role, but it should be first bounded by energetic constraints. We also plan to incorporate environmental noise, that is to say, incorporate in the model the realistic assumption that the environment changes through time and therefore so will the fitness of different species in the meta-community. The ultimate goal is to provide a Unified Theory that can make clear predictions about size-abundance-distributions in natural systems, and, perhaps, also make predictions about speciation-extinction dynamics. We currently hold data to test predictions on an 'ecological' time scale. At this point we are uncertain on whether we could obtain good fossil/contemporary data to calibrate/test models.

Interested:
Diego Barneche Rosado (diego.barneche@mq.edu.au)
Cornelia Metzig (cornelia.metzig@hotmail.com)
Michael Kalyuzhny (michael.kalyuzhny at mail.huji.ac.il)
Ana María Gómez López (anamaria.gomezlopez@yale.edu)
Sean Gibbons (sgibbons at uchicago dot edu)

Coupling of different types of networks

Coupling of different types of networks

We want to develop a concept of how to work with networks of qualitatively different types of relationships or interactions that can influence each other (eg. natural and social). Sign up and come to brainstorm with us :) Contact: sanjakojasanja@gmail.com

For example, people's beliefs about the health benefits/risks of vaccination can be influenced by their social network, and may be studied using belief propagation models. Simultaneously, diseases may spread through a population, which can be studied using diffusion or other epidemiological models. Furthermore, people's beliefs about vaccination may affect their probability of getting infected by a disease, and in turn, getting infected may cause them to re-evaluate their beliefs.

Another example could involve an ecological network expressing predation and competition among species coupled with an environmental network.

Interested:
Sanja
Fahad
Brian
Nhat (nhattdnguyen@gmail.com)
Beth Lusczek
Glen Otero (gotero@linuxprophet.com)
Hiroshi (hashika1@jhmi.edu)

Please fill the doodle
http://doodle.com/m99hkws4k46icx8b

Co-evolution of Anti-vaccination Sentiment and Flu Infections

People's beliefs about the health benefits/risks of vaccination can be influenced by their social network, and may be studied using belief propagation models. Simultaneously, diseases may spread through a population, which can be studied using diffusion or other epidemiological models. Furthermore, people's beliefs about vaccination may affect their probability of getting infected by a disease, and in turn, getting infected may cause them to re-evaluate their beliefs.

Previous work has separately studied (1) how beliefs propagate and change over time, (2) how diseases spread through a population over time, and (3) the (static) correlation between beliefs about vaccination and infection rates; but possibly not all three simultaneously.

This project has some similarities to the project on "Coupling Different Types of Networks," and it may make sense for the two groups to be in communication with each other, or even to merge the two projects. This can be discussed later as the projects develop.

Interested:
Ells
Nhat
Glen
Andy
Brian
Cecilia (ci.andreazzi@gmail.com)

Consciousness as an emergent state of matter – what do you think?

Emergence of consciousness
You’re conscious right now, reading this. How does subjective experience emerge out of the bundles of particles that we all are?

Scholars of many fields have been deconstructing the mind/body dualism for a while, but consciousness remains a big, hard question. I’m no expert and by proposing this as a topic I’m not expecting that any of us will solve it, but I would be very interested in exchanging on the issue with the smart individuals that you are, grounded in so many backgrounds and unafraid of complex problems.

So... physicians and physicists, social scientists, biologists, mathematicians, philosophers, computer scientists and others… what do you think? Don’t hold back – if beer is necessary for you to address this issue, it can be arranged.

Contact Sarah L (laborde.7@osu.edu)

Interested:
Alireza (alireza.goudarzi at gmail.com)[totally agree with writing down something to figure out what are plausible ways to think about and study this]
Fahad (fahad.khalid@hpi.uni-potsdam.de) [love the topic ... I'm glad someone brought it up ... I might have some ideas to contribute]
Claire (lagesse.claire at gmail.com) [it sounds fascinating... :) ]
Stefan (s.pfenninger12@imperial.ac.uk)
Hiroshi (hashika1@jhmi.edu) [Happy to contribute from a medical and personal point of view]
Sean Hayes(shaye004@ucr.edu)
Cole Mathis (cole.mathis@asu.edu) [The origin and consciousness and the origin of life (my main research interest) have a lot in common, I'm always down to talk about consciousness with some beers, if something novel emerges that's great.]
Emília [happy to contribute from various points of view, but I advance that, to me, it has a lot to do with memory]
Ana María (anamaria.gomezlopez@yale.edu)
Brian (bthompso8784@gmail.com)
Matthew Ayres (matthew.ayres@growthandinnovation.com.au)
Beth Lusczek (lusc0006@umn.edu)
Stojan (stojand@mpib-berlin.mpg.de) (sounds perfect for a beer discussion)
José Aguilar-Rodríguez (jose.aguilar@ieu.uzh.ch)
Bernardo Furtado (bernardo.furtado@ipea.gov.br)(Happy to learn. Consciously on subconsciously.)
Nicolas K. Scholtes (nicolas dot scholtes at unamur dot be) [The science fiction aficionado in me is jumping up and down in his chair!]
Renske Vroomans (R.M.A.Vroomans@uu.nl) [I'd love to join the conversation, I love big topics!] Tom McAndrew [Thomas.McAndrew@uvm.edu]
Diego Barneche (digo.barneche@mq.edu.au)[glad to join the conversation over beers and maybe formalizing a project]

Note: this doesn’t have to become a formal project, although it could. Let me know if you’re interested in a chat, writing an interdisciplinary dialogue piece, or anything related to this question.

  • There is a published paper written by a MIT researcher "Consciousness as a State of Matter", just in case if you do not know this... Ernest
  • Thanks Ernest! I have a copy of this paper and I added a link on the webpage: http://tuvalu.santafe.edu/events/workshops/index.php/Emergence_of_Consciousness_Page
  • As far as meeting, most people can meet tonight except Ana-Maria and myself (going to the opening of this arts festival: http://currentsnewmedia.org ).. otherwise it looks like sunday over lunch (in the cafeteria) is the go. It's so many of us, there will no doubt be many partial meetings, so let's keep updating Alireza's page! Yesterday Emilia and I had a chat about this with Alfred Hubler about this, we will add ideas to the page later today. Cheers! Sarah

Tradeoffs between division of labor and stability in networks

The Black Queen Hypothesis BQH describes the evolution of functional dependencies in microbial ecosystems. This process results in a subset of a community providing necessary services for the rest of the community (see link). Organisms that can outsource essential functions escape the cost performing these functions and have more resources for growth and reproduction. However, this process makes the ecosystem more delicate, as the destruction of key species can eliminate their crucial service(s) and lead to system collapse. Thus, there is a tradeoff between the stability of a network (in the face of perturbations), and the degree of cooperation (how many tasks can be outsourced). As a result of this tradeoff, we could expect different community types to arise in stable vs. variable environments. There is likely some critical range between these two modes (high vs. low environmental variability), where some mixed strategy is optimal.

This process may have an analogue in the development of multicellular life (vs. free-living single-celled organisms), where each cell type expresses a subset of the genome and provide a specific set of services to the whole (division of labor). This might also be reflected in social or economic networks (higher stability = more cooperative?). A connection could potentially be made to life-history tradeoffs for individual organisms (r vs. k selection - oligotrophs vs. copiotrophs), or in dissipative chemical systems (e.g. Stat-Mech of Self Replication).

Definitely a work in progress, please add your thoughts if you are interested! Also, please include your contact info alongside your name.

NEXT MEETING: Thursday at 2PM - coffee shop

  • Thoughts:

Vipin: I believe that this mechanism may have much to do with why "business cycle" occur.

Luis: Maybe this could be also related with the structure and complexity of institutions and its stability. Successful societies increase their size and develop more complex institutions (with higher level of bureaucracy for example) and that can make them less "flexible" and susceptible to get destroyed under not predicted crisis. I think this is related with "social entropy"...

Ali: This paper might be of interest: http://people.biology.ufl.edu/ulan/pubs/Zorach.pdf They use Shannon's entropy to calculate "number of roles" or division of labor in a network. It is related to trade-off of division of labor and robustness.

Rohan: Potential evolutionary network model: N different kinds of edges for N resources (so N separate networks). Directed graph, with self-loop signifying self-sufficiency for a particular resource, and a directed edge from A to B indicates that A relies on B for a particular resource. The weight of an edge corresponds to the proportion the population of node A relying on node B for that resource. Evolution: at each time step, random mutations change the weights of the edges, and fitness calculations based on the number of self-produced resources change the composition of each node over time. A perturbation would involve randomly converting all the edges of a particular node to self-loops (or possibly just deleting the node).

Ernest: a agent-based model: Let’s imagine that there is a land where resources A, B, C are located, and a kind of creature which need all of these resources to survive (imagine many creatures live in this land). One surviving strategy is that every creature gathers A, B and C and eat on its own. But maybe under some condition, specialization would automatically happen (e.g. one creature gather A and another gather B, and then they share). The goal is to find the necessary condition (we could think out of many sufficient conditions, but finding necessary conditions seems not so easy). PS: My original plan is to investigate the process before labor-division, that is, define some creatures' behavior rules which would not directly tell them to collaborate, to see whether they would figure out some way to collaborate automatically. Of course, this kind of rules should satisfy some conditions, otherwise labor-division would not happen definitely. So these conditions matter, and it's what I want to investigate. Anyway, my future plan is flexible.

George: I am thinking of the dependency within the financial networks which you can see cascading failures in a lot of past crisis. i am interested in using the model Rohan mentioned above by adding some some mechanism like internal effect.

Sean: Here are some resources for modeling dynamical networks in python pycx

  • Interested:

Sean Gibbons (sgibbons at uchicago dot edu)
Rohan Mehta (rsmehta at stanford dot edu)
Stojan (stojand@mpib-berlin.mpib.de)
Emília (emilia.garcia.casademont at gmail dot com)
Ali Kharrazi ali[at}pp dot u-tokyo dot ac dot jp
Vipin (vipin.veetil at gmail dot com)
Ernest Liu (yu.liu@math.uu.se)
Cole Mathis (cole.mathis@asu.edu)
José Aguilar-Rodríguez (jose.aguilar@ieu.uzh.ch)
Renske Vroomans (R.M.A.Vroomans@gmail.com)
George (qiaozhi827@gmail.com)

How can evolutionary game theory be applied to electricity trading

The interest of this project is the following: how can evolutionary game theory be used to find optimal strategies for consumers and/or producers that bid in the electricity market. Also some real data can be considered.
Contact Blaz

Interested:
Alberto
Leo
Ali ali [at] pp.u-tokyo.ac.jp

The project is currently not active. Due to the fact that I am just starting this research, I still don't have a project-appropriate well-defined idea on which we could work on. Still, it might happen that we can start the project towards the end of the school and then continue to work on it after the school is ended. Thank you all so much for your interest in the project. I wish you best luck with your other projects.

Information Theory of the Heart

If you want to discuss more please contact Hiroshi (hashika1@jhmi.edu).

Interested:
Hiroshi (hashika1@jhmi.edu)
Brian (bthompso8784@gmail.com)
Flavia (flamarquitti@gmail.com)
Beth (lusc0006@umn.edu)
José (jose.aguilar@ieu.uzh.ch)
Degang (samuelandjw@gmail.com)
Shai (shai.gorsky@utah.edu)
Nix (nix@math.ucdavis.edu)
Josh (garland.joshua@gmail.com)

Cells and Software: Is Evolution a Software Engineer?

There appear to be striking similarities between how we design software, and how evolution designed cells. Some of the analogies are:
- The concept of "Encapsulation". In object oriented programs, data inside an object is protected by an interface of functions. Similarly, processes within a cell (intra-cellular signaling cascades) are protected from the extra-cellular messaging activities through membrane-bound receptors acting as the interface.
- Apoptosis (programmed cell death) is similar to proper memory deallocation in programs, while Necrosis is similar to dangling pointers and memory corruption.
- Proteolysis (breaking down of proteins into constituent amino acids) is similar to automatic garbage collection.

There are certain constraints (perhaps physical constraints) on the available solutions in biology. Solutions within this constrained solution space eventually result in the emergence of complex behavior. This emergent behavior has enabled humans to engineer solutions to everyday problems. These artificially engineered solutions are very similar in principal to the solutions that already exist in biological systems.

Are our creative processes ultimately bound by the physical constraints that underlie molecular mechanisms? Or, do we just tend to interpret the phenomena at the molecular level according to our own understanding, which is limited by our senses and neurological processes? Is it all about optimization, and everything else is just a side effect? Can/do “Patterns” transcend disciplines?

Motifs (biological systems) – Patterns (software). Patterns/motifs transcending biological and software systems could perhaps, fill gaps in our knowledge of biological systems, and help us design better software systems. Deterministic patterns could perhaps indicate what is required for high level functions to emerge from molecular interactions. And perhaps these similarities occur at the interface between low level interaction and emergent phenomena.

So let's further explore biological and software systems, and try to find answers to these questions.

For more information, please contact Fahad (fahad.khalid@hpi.uni-potsdam.de).

Interested:
Fahad
Ernest (yu.liu@math.uu.se)
Diana (dianalg11@gmail.com)
Degang (dwuab@ust.hk)
Stojan
Ana María (anamaria.gomezlopez@yale.edu)
Brian (bthompso8784@gmail.com)
George (qiaozhi827@gmail.com)
Glen (gotero@linuxprophet.com)
Renske

Susceptibility of Fields of Research to Interdisciplinary Influences (network perspective)

Interdisciplinary research networks

Fields of Research are more often than not isolated from one another in terms of their community, jargon, perspective, research programme (in the sense of Lakatos) and their journals and lots more.

This isolation stands in the way of a fruitful merging and interaction of fields. One needs to understand the obstructions. Hence an analysis of the determining factors for 'community inbreeding' is sought after. Eventually also a quantitative measure for the susceptibility of a research field could be formulated.

A clear-cut project could be based on data together with a network type analysis of a precise question around the determining factors. We could look at citation data or journal based data or maybe there is a chance to get hold of data from Jessica's collegues, who works on the diffusion of knowledge.

It would be great also to discuss and gather our ideas and experiences.

Meeting: June 12, 4:20, balcony outside Great Hall.

Interested
Leo (horstmey@mis.mpg.de )
Jennifer Hellmann
Anna Olson
Lin Li
Stefan Pfenninger [other angles include: citation networks/academic generations and how they change through time (Diego's idea); citations between subfields to quantify interdisciplinarity and how it changes over time]
Catherine Bale
Sarah Laborde [laborde.7@osu.edu - sorry I probably missed a lot of the initial discussion.. but I would love to listen to/participate in the next meeting]

@ Diego's idea, I remembered this quote by Max Planck: “A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.”

Simple case for studying about “from single-cell to multi-cell / species collaboration”

QUESTION: how single-cell evolves to multi-cell, even they do not have so-called “intelligence”?

How to do: Let’s imagine that there is a land where resources A, B, C are located, and a kind of creature which need all of these resources to survive (imagine many creatures live in this land).
One surviving strategy is that every creature gathers A, B and C and eat them on its own. But maybe under some condition, specialization would automatically happen (e.g. one creature gather A and another gather B, and then they share).
The goal is to find the necessary condition if possible (we could think out of many sufficient conditions, but finding necessary conditions seems not so easy).

Interested:
Ernest Liu (yu.liu@math.uu.se)

comment: are you familiar with holland's "echo" stuff ..stojan

COMBINED with project 17 "Tradeoffs between division of labor and stability in networks".

An Analysis of the Hierarchies Present in Modern Economic Theory

The idea of this project is to provide a comprehensive survey of the hierarchical structure of economics (macro -> microeconomics) following the reductionist = constructionist + emergence framework of Anderson (N.B. This equation is my own interpretation of Anderson's paper and is open for discussion). Ideally, we would first investigate the microfoundations literature of macroeconomics and the possibility for ABM to rectify some of the redundancies therein after which we would reverse the direction and see how, given some 'fundamental' laws of economic behaviour, we could couple these with emergent phenomena (whatever they may be) to reconstruct the economy at the macro-level.

Interested
Nicolas ( nicolas dot scholtes at unamur dot be)
Claudius
stojan (stojand at mpib dash berlin dot mpg dot de) i'm happy to discuss this with you guys..and hear your opinion

The Topological Modeling of Infrastructure Networks

There is some disagreement on the topology of power networks in the literature. There are preferential attachment, small-world and random graph models proposed to explain the topology of power networks. There are caveats associated with each of these generative models. For instance, there is the cost of building new transmission lines that may question the validity of preferential models.

The broad question that we attempt to answer here is:

-Can these previous network models create real-world power networks?
-Can we come up with a better model that can replicate a power network or at least one with similar characteristics?
-What about other infrastructure networks, e.g., natural gas network, water piping networks, etc?

Interested: Pooya Rezaei
Tom McAndrew

Functional Networks and their evolution

We are interested in researching (possibly hierarchical) networks that have a function which can be quantified. We are interested in understanding how such networks evolve and may continue to deliver all or part of their function when they are damaged. As of now, we do not have a clear conceptualization of these ideas in formal network language, and our immediate task is to think of the formal network structure we wish to explore (e.g. b-partite, multileveled). Possible concrete cases to apply this formal characterization to are trees (i.e. plants), power grids, or financial networks.

Interested (please contact us to join the debate):

Alex Brummer (brummera@email.arizona.edu)
Shai Gorsky (shai.gorsky@utah.edu)
Beth Lusczek (lusc0006@umn.edu)
Stefan Pfenninger (s.pfenninger12@imperial.ac.uk)
Jessica Santana (jsant@stanford.edu)
Nicolas K. Scholtes (nicolas dot scholtes at unamur dot be
Ali Kharrazi
Renske Vroomans (R.M.A.Vroomans@uu.nl)

Complex systems approaches applied to data scarce environments - Logone floodplain of Cameroon

Field-based research, be it qualitative or quantitative, often yields fragmentary and limited data sets when collected in settings with political instability, economic disparity, and rapidly changing environmental conditions. Using the Logone river floodplain in Cameroon as a case study, this project aims to carry out multi-scalar analysis of fragmentary data sets to understand larger social dynamics and ecological regime shifts in this region. We want to experiment with the benefits and limits of complex systems approaches and tools (starting with a common pool resources model) when applied to a complex case study.

Next meeting 2:45 Tuesday in lecture room (to start with).

Sarah L
Ana María
Claudius
Luis Martínez (fnxabraxas@gmail.com)
Ali Kharrazi
Francesca Lipari
Brian (bthompso8784@gmail.com)

The architecture of an empirical genotype-phenotype mapping

One of the central open questions in biology is to understand how genotypes map onto phenotypes. Our current knowledge on this question comes from computational models that allow us to rapidly map genotypes to phenotypes for some biological systems, facilitating the systematic exploration of their vast genotype spaces. These models have shown that many genotypes usually map onto the same phenotype. These genotypes form genotype networks, or neutral networks, that spread far away into genotype space. These networks contain as vertices all the genotypes that share a same phenotype, where genotypes are connected by edges if they differ by a single mutation.

Thanks to several computational models, the formalism of genotype networks has been very successful at providing new insights about the evolution of systems as diverse as RNA, proteins, regulatory networks and metabolism. Payne & Wagner have recently pioneered the application of this formalism to transcription factor (TF) binding sites. TFs are DNA-binding proteins that regulate gene expression by binding to short sequences on DNA — TF binding sites — that are in close physical proximity to the genes’ coding sequence, thus inducing or repressing gene transcription. The set of DNA sites bound by a particular TF can be viewed as a genotype network. This mapping from TF binding sites onto their cognate TFs constitutes the first exhaustive genotype-phenotype map entirely based on experimental data.

Research questions that can be asked in this system:
1. What is the geometry of empirical genotype networks in genotype space and how that geometry affects evolutionary exploration?
2. What is the community structure of these empirical genotype networks? Is this structure determined by the biophysics of TF-DNA binding?
3. How accesible is one phenotype from any other phenotype in this system and how that accessibility defines phenotype space as a topological space?

If you are interested please contact José Aguilar-Rodríguez (jose.aguilar@ieu.uzh.ch).

References:
1. Stadler, B.M., Stadler, P.F., Wagner, G.P., and Fontana, W. (2001). The topology of the possible: formal spaces underlying patterns of evolutionary change. J. Theor. Biol. 213, 241–74.
2. Payne, J.L., and Wagner, A. (2014). The robustness and evolvability of transcription factor binding sites. Science 343, 875–7.

Interested:
José Aguilar-Rodríguez
Cole Mathis
Renske Vroomans
Leto Peel
Massimo Stella

Glen Otero (gotero@linuxprophet.com)

Leo Horstmeyer (horstmey@mis.mpg.de)

Brian Thompson (bthompso8784@gmail.com)

Sean G. (sgibbons@uchicago.edu)
stojan (stojand@mpib-berlin.mpg.de)

Network robustness as a function of nodes' resilience

Are there characteristics of a network's nodes that make the network more or less robust to failure? Perhaps degree assortativity, or nodes' positions relative to the demands made on them? Given that network failure can affect every node in a network, nodes might, as it were, gain an advantage from being positioned far away from nodes that could catastrophically fail.

Interested:
Anna Olson (olsona at cs.uchicago.edu)
Jenn Hellmann
Alireza Goudarzi (alireza.goudarzi at gmail dot com)
Tom McAndrew

Optimal Quarantine Strategies in Financial Networks

We are interested in borrowing and adapting concepts from epidemiology to provide policy guidelines that minimize the damage to the global financial system when defaults cascade between financial institutions.

From a network perspective, we would like to identify optimal quarantine strategies (selective edge pruning, node isolation, community isolation, etc) by simulating default cascades of variable intensity on an idealized financial network. A key component of this analysis would focus on the temporal delay between the identification of a cascade threat and implementation of a control measure.

Interested:

Cough analysis

This project will investigate if signal analysis of a person's cough coupled with collection of other data (such as: age, height, weight, etc.) can be used to distinguish between a healthy and an unhealthy lung. Our idea is that humans have evolved a specific sound/frequency of coughing so as to break up mucus in the lungs and that diseased lungs will have different sounds/frequencies than a healthy lung. Using the collected data, we will be able to determine if coughs are specific to individuals or if there are qualities about them that span a population. If there are qualities that span a population, then the data can lead to answering deeper questions about how to care for individuals with cystic fibrosis. For instance, treatment of this disease state could lead to adjusting their cough—through external means—to help move stagnant mucus in their lungs.

Project goals

  • Create a website that allows users to submit cough sound files and enter appropriate metadata to.
  • Determine relevant signal analysis techniques that can be used to analyze coughs.
  • Create a classifier that can separate coughs into different categories.
  • Display any analysis on the website.

We will be using standard techniques to create the website (Ruby on Rails), host the data (Postgres and Amazon S3), generate a classifier (Mahout), and display the data (D3). If you are interested in learning some of the standard technologies that companies use, you are welcome to join the group. The project is engineering heavy, however, the scientific goal is to determine if a cough is either specific to a person or if there are principles about it that are similar within a population. If there are similarities, then this can lead to interesting biological questions such as how did the cough evolve.

Interested

  • Andy Email Andy using the wiki if you are interested. amaloney theatthingy austin.utexas.edu
  • Ells
  • Nix has a nasty chest cold and would be interested in learning about some of the software used here.
  • Glen
  • Pooya (pooya.rezaei@uvm.edu)
  • Renske -would be interested in learning the techniques, probably can't contribute much else

Using Natural Network Lessons in Modern Network Environments

There is a significant set of natural and biological network data that has been investigated. How can insights, lessons and parallels be applied to modern organisations, modern networks and modern contexts? Can innovation, idea diffusion and increased resilience be informed by key insights?

This group has met twice to form initial views and if there is enough interest (>3-4 people) will continue or if not may merge with other innovation and network related areas.


Group Members

Matthew Ayres (matthew.ayres@growthandinnnovtion.com.au)
{members please add names!}

Interested

Diego Barneche (diego.barneche@mq.edu.au)[glad to provide some insights in terms of ecological resilience and learn of possible parallels in modern organisations]

As the original 5 members have joined similar groups in Ecology, Innovation and Resilience - this group will not be further pursued.

Economic inequality, kinship networks, and political transitions

Economic inequality, kinship networks, and political transitions project page

Summary:
It has been argued that political institutions are central to long-term economic development. Acemoglu and Robinson (2006) suggested that economic factors, namely the distribution of wealth, play a central role in political transitions (e.g. transition from non-democratic to democratic regimes). Building on Demetrius and Manke (2004), we would like to additionally consider the role of kinship networks in political transitions. That is, for a given state of the economy and distribution of wealth, what role do kinship networks play in facilitating or precluding institutional change? How might the outcome change with an altered distribution of wealth? As such, we implement a simple agent-based model to gain insight into the institutional outcomes of select historic events and, if possible, inform policy-making.

Group members:
Bernardo
Catherine
James G.
Heath (hendersonhl@gmail.com)
Leo
Brais

Chocolate and Coffee Trade as a Power Grid

Motivation:
Previous research on global resource trade networks has mostly viewed trade flows as a static state. Global resource trade networks however are dynamic and are responsive to fluctuations and perturbations. As a new research approach this group intends to explore common properties between global resource trade networks as an electric power grid. We will strike analogies between the trade flows of chocolate and coffee and power grid dynamic measures. Specifically, we intend to describe an chocolate and coffee trade systems in terms of voltage, current, capacitor, resistance, and inductance.

The resulting dynamic matrix of chocolate and coffee trade will allow us to make delicious evaluations of the resiliency of these trade networks. Insightful questions can be approached, such as: how responsive are these networks to perturbations and disruptions - that is, are we in danger of chocolate/coffee blackouts ... !

Group members:
Tom
Pooya
Junjian
Ali
Stefan
Nicolas