Osion kuvaus

  • Project: Animal social networks
    Data description: This project will use data from the Animal Social Networks Repository, which is a fairly new public repository that contains networks representing the social interactions between many animal species, from ants to primates. Different networks represent different interactions, from physical proximity to grooming.
    Possible research questions:
    • How do the properties of social networks of different species reflect their position in the tree of life / taxonomy?
    • How close are primate social networks to those of humans?
    • Which animals have the most complex networks (largest deviations from randomness)?
    (See also this paper and the associated data set: https://www.biorxiv.org/content/10.1101/2020.01.17.910695v1.full)

    Project: Brain networks
    Data description: Two preprocessed datasets (Autism Brain Imaging Data Exchange, ABIDE and UCLA Consortium for Neuropsychiatric Phenomics, CNP), containing anonymized MRI and rsfMRI images and phenotypic information from 138 individuals. The subjects are classified as typical healthy (or "typical") controls (TC) or diagnosed with: Autism Spectrum Disorder (ASD), bipolar disorder, schizophrenia, and attention-deficit/hyperactivity disorder (ADHD).
    Possible research questions:
    • Is brain connectivity altered in people with neuropsychiatric illnesses?
    • Are some brain networks similar in people with different neuropsychiatric disorders?
    • What is the relationship between abnormal brain connectivity and the symptoms/associated features of neuropsychiatric disorders?

    Project: Board game networks
    Data description: Many board games, such as Pandemic, operate as network processes. In the best of these games, the network topology and process defines the gameplay. I have created an open dataset of various board game networks for your enjoyment.  In this project, you will understand the board game networks, understand the games, and then interpret the network in the context of the game. The project instructor also provides mentoring in open data and scientific software development.
    Possible research questions:
    • How do the networks affect gameplay, compared to random (planar) graphs?
    • Can networks of different games be distinguished from each other?  Is this by design?
    • Can you make a "network scientist's cheatsheet" for each game which can be used to guide player strategy?

    Project: Erasmus mobility network
    Data description: The project will use data from the Erasmus programme, which is one of the biggest collaboration networks in Europe. This data set contains the flows of students, teachers, and administrative staff between institutes, regions, and countries, as well as data about the points of interest (POI) in the neighbourhood of the institutes. Directed weighted network s or multi-layer networks would be ideal for representing the data.
    Possible research questions:
    • How to model the education-related mobility as a network? Is it similar to immigrant or commuting mobility?
    • How to determine the rank of institutes based on the mobility network? Are the ranks based on students’ flow different from the ranks based on the flows of teachers and staff?
    • How to model the mobility based on different majors (humanities, engineering etc.) separately?
    Link to data descriptor: https://www.nature.com/articles/s41597-020-0382-1
    Link to data: https://data.mendeley.com/datasets/vnxdvh6998/3
    (See also this paper using this dataset: Investigating collaborative and mobility networks: reflections on the core missions of universities. https://link.springer.com/article/10.1007/s11192-021-03865-7
    and the further optional reading materials:
    Maximilian Schich et al. 2014. A network framework of cultural history. doi.org/10.1126/science.1240064
    Aaron Clauset et al. 2015. Systematic inequality and hierarchy in faculty hiring networks. doi.org/10.1126/sciadv.1400005
    Jaehyuk Park et al. 2019. Global labor flow network reveals the hierarchical organization and dynamics of geo-industrial clusters. doi.org/10.1038/s41467-019-11380-w)

    Project: Robustness & resilience of transport networks under extreme weather conditions
    Data description: Finnish Transport and Communications Agency provides long-term historical traffic road data (e.g. traffic volume, traffic incidents, weather, (ice) maintenance trackings, etc.). In this project, you can explore the conceptions such as robustness, resilience, percolation. Notice: This project requires you to build a data set from scratch, but the project instructor will provide mentoring in data scraping and preprocessing.
    Possible research questions:
    • How to measure the robustness of a transport network? Is it consistent with empirical data?
    • How to model the resilience of a transport network with empirical data?
    • Extreme weather is difficult to predict. How to improve the robustness of the transport network under the uncertainty of extreme weather?
    Link to data: https://liikennetilanne.fintraffic.fi/
    Link to data API: https://www.digitraffic.fi/en/road-traffic/#restjson—apis
    Tool for constructing road network: https://github.com/gboeing/osmnx
    (As optional reading material, see also:
    Wang, W., Yang, S., Stanley, H. E., & Gao, J. (2019). Local floods induce large-scale abrupt failures of road networks. https://www.nature.com/articles/s41467-019-10063-w)

    Project: Social networks of students
    Data description: A multi-layer temporal network which connects a population of more than 700 university students over a period of four weeks. The dataset was collected via smartphones as part of the Copenhagen Networks Study. Included: the network of physical proximity among the participants (estimated via Bluetooth signal strength), the network of phone calls (start time, duration, no content), the network of text messages (time of message, no content), and information about Facebook friendships.
    Possible research questions:
    • How are the different layers of the multi-layer social network related (proximity, FB, calls, text messages)?
    • Is it possible to learn and predict behavioural patterns of individual students?
    • How do the networks evolve in time, from the beginning of the study to the end?