Social System Inference From Noisy Observations

This article studies social system inference from a single noisy trajectory of public evolving opinions, wherein observation noise leads to the statistical dependence samples on time and coordinates. We first propose cyber-social that comprises individuals in network set information sources cyber layer, whose opinion dynamics explicitly takes asymmetric cognitive bias including confirmation negativity process into account. Based proposed model, we then study sample complexity least-square auto-regressive model estimation, which governs length observed is sufficient for identified achieve prescribed levels accuracy confidence (PAC). Building investigate inference, with particular focus weighted topology parameters bias. Finally, theoretical results effectiveness framework are validated by U.S. Senate Member Ideology data.