Research

Research by NICO faculty is funded by the National Science Foundation, the Army Research Lab, the Office of Naval Research, the Air Force Research Lab, the National Institutes of Health, the Howard Hughes Medical Institute, the MacArthur Foundation, the McDonnell Foundation, DARPA, and many others.  Current topics include

  • Structure, dynamics, evolution, and control of complex networks
  • Big Data, data science, analytics (See also: https://datascience.northwestern.edu/)
  • Robotic sensor swarms
  • Group decision making in online environments
  • Contagion on social networks and viral marketing
  • Cell-signaling networks as a tool to understand human disease
  • Transportation networks and disease epidemics
  • Elections and voting patterns
  • Stock trading and prediction markets
  • Logistics and supply networks
  • Network models of cancer
  • Semantic networks and evolution of language
  • Agent-based modeling
  • Science of team science:  developing tools supporting the formation of effective interdisciplinary research teams

Some of these topics and others are highlighted in greater detail below.

neurons

Computational Neuroscience

NICO research in computational neuroscience includes developing computational models of principal neurons and interneurons in the hippocampus as well as the neural circuitry of the mammalian retina and pathways responsible for night vision (rods) and daylight vision (cones).

 
cancer Epigenetic Memory and Cancer Development

Increasing evidence indicates that cancer results not only from changes in genetic information, such as point mutations, chromosomal rearrangements, and gene segment amplification and deletion, but also from changes in epigenetic information. Recent theoretical analyses of epigenetic regulation in a model organism (yeast) show that ideas from the dynamical systems field may explain the stability of epigenetic states and the dynamics of spontaneous changes between them. NICO researchers are developing quantitative predictive models of epigenetic inheritance and switching in these systems.

 
NetLogo Agent-Based Modeling

Uri Wilensky's Center for Connected Learning leads the development of NetLogo, popular freely avaiable software for simulation and interactive experimentation with multi-agent systems.  NetLogo allows easy creation of agent-based models for research and education.

 
swarm robots Robotic Sensor Swarms

Mobile sensor networks can be deployed for tasks such as environmental monitoring, surveillance, or searching a collapsed building for survivors. Decentralized control algorithms allow the mobile sensors to adapt to a changing environment and to failure of individual sensors, without the need for human intervention or a centralized controller. NICO faculty are developing control theory to support "swarms" of mobile sensors, including tools to "compile" desired emergent behaviors into control laws and estimators for individual robots that interact over a wireless communication network.

 
air network Transportation Networks

Large-scale disasters that interfere with worldwide air-travel trigger immense socio-economic costs. Yet their impact is poorly understood and difficult to quantify. In this project we characterize and quantify the impact of these disasters and determine if there are universal features across events. Examples are the eruption of volcano Eyjafjallajökull, the September 11th terrorist attacks and geographical disruptions in general. We find that the susceptibility of airports to all geographical disruptions follows similar, strongly heterogeneous distributions that lack a scale. On the other hand, airports are more uniformly susceptible to attacks that target the most important hubs in the network, exhibiting a well-defined scale. We find that the same mechanisms responsible for efficient passenger flow may also keep the system in a vulnerable state.

 
social network Interaction between Social and Information Networks

NICO faculty are examining how information in the public domain jumps into the private information domain--what people talk about in their private chats and instant messages. It is currently unknown which properties of public domain information make it likely to jump into the private communication domain where select decision makers process it. This jump indicates selective value and relevance for action. This research also considers the properties of private information that induces action; in other words, not just what makes a meme popular, but what makes it influential in changing behavior.

 
team science The Science of Teams

We now know that scientific discovery is spurred on when diverse ideas are united or when proven innovations in one domain are introduced into a new domain, inspiring fresh thinking about old problems. These structural preconditions suggest that most scientific advancement is the consequence of complex networks of individuals, and not the inspired insight of loners. Our research has demonstrated that high impact science is increasingly a product of teams and networks. Across all areas of science, teams on average produce more highly cited papers and patents than individuals, a gap that is widening annually. To continue the spectacular growth rate of science and technology, we aim to provide scientists and practitioners with advanced research for strategically navigating and harnessing the networks of knowledge and scientists worldwide.

 
disciplines The Emergence of Scientific Disciplines

The SONIC research group is investigating the emergence of scientific disciplines. Researchers are using statistical and computational models to analyze how the collaboration networks of medical specialists, scientists, and scholars explain the emergence of Oncofertility as a new subdiscipline at the intersection of oncology and reproductive science.

 
traffic signal Urban Traffic Signal Timing

The problem of automotive traffic flow on freeways has received extensive study by physicists and applied mathematicians. However, the movement of cars in dense urban grids controlled by traffic lights has not yet received the same level of attention from a theoretical point of view. What is known is that a "green wave" can provide ideal traffic flow on a major street in a single direction, but that it's impossible to arrange green waves in multiple directions simultaneously. This project's goal is to better understand the mathematical structure of the problem underlying the timing of traffic lights, and to explore conditions under which "frustration" in the system might be minimized, thus allowing for better overall efficiency.

 
weight loss

Weight Loss in the Internet Age

Health care costs due to weight-related complications continue to rise. It is important to identify the weight loss strategies most likely to lead to individual success, and these strategies must be scalable to millions of people. Successful weight loss programs involve regular interaction with trained medical personnel, but this sort of interaction is too costly to provide to millions of individuals. What other options are there? A new strategy brought about by the blossoming of the internet is an online weight management program. These web sites and phone apps allow their users to not only track their weight, their calorie intake and daily activity, but even discuss topics of interest in blogs or forums, and make friends with other users in the system. NICO researchers are studying the correlations between weight loss and individual characteristics (such as age, initial weight, level of online activity) and network features (such as degree, clustering or community membership).