Analysis of network evolution and learning

Abstract

Much of society is organized in networks and networks are important because individuals typically interact largely or perhaps entirely with those to whom they are closest in the network (not necessarily physically). The questions which this research addresses are: How do networks evolve? What networks emerge? How does information about the individuals in the network affect the evolution and the eventual shape of the network and conversely, how does the evolution of the network affect learning about the individuals in the network? How do the evolution and the shape of the network affect the functioning of the network (e.g. productivity, social welfare, stability, fragility etc.)? What social norms promote the formation and maintenance of well-functioning and robust (and stable) networks?This project addresses these questions in the context of information networks, where the underlying activity is theproduction, consumption and dissemination of information. But the lessons learned will surely apply to other networks as well.An essential part of the research agenda is driven by the understanding that individuals are not all the same - some are better producers of information, some are better disseminators of information, some have different values for the consumption of information - and the characteristics of an individual may change over time. These characteristics are not known in advance but must be learned and perhaps relearned. Moreover, it is not only the intrinsic characteristics of the individuals that are important, but also their connections (i.e. the links they have already formed).

Document Details

Document Type
DoD Grant Award
Publication Date
Jan 04, 2017
Source ID
N000141712125

Entities

People

  • Mihaela Van Der Schaar

Organizations

  • Office of Naval Research
  • United States Navy
  • University of Oxford

Tags

Fields of Study

  • Computer science

Readers

  • Economics
  • Neural Network Machine Learning.
  • Theoretical Analysis.