CEU Electronic Theses and Dissertations, 2007
| Author | Tayanovskyy, Anton |
|---|---|
| Title | Agent-based modeling of cascades on social networks |
| Summary | Patterns of conformity in agent behavior, known as cascades, are observed in many areas including technology diffusion, evolution of social norms, fashion, politics, and multimedia. These patterns are consistently found to be unpredictable and fragile. Traditional explanations in the economics literature focus on the micro-level and include rational herding and observational learning. In contrast, statistical mechanics has a macro-level methodology that deals with similar phenomena. This paper seeks to integrate the two approaches. Cellular automata are adjusted to become closer to real social networks in their topology and equipped with a probabilistic update rule compatible with economic cascade models. The resulting flexible agent-based framework is used to simulate cascades and reproduce their key features, including fragility. |
| Supervisor | Medvegyev, Peter |
| Department | Economics MA |
| Full text | https://www.etd.ceu.edu/2007/c05taa01.pdf |
Visit the CEU Library.
© 2007-2021, Central European University