A1162
Title: A random graph-based autoregressive model for networked time series
Authors: Weichi Wu - Tsinghua University (China) [presenting]
Chenlei Leng - University of Warwick (United Kingdom)
Abstract: Contemporary time series data often feature objects connected by a social network, which naturally induces temporal dependence among connected neighbors. The network vector autoregressive model is useful for describing the influence of linked neighbors, while its recent generalizations aim to separate influence and homophily. Existing approaches, however, require either correct specification of a time series model, accurate estimation of a network model, or both, and rely exclusively on least squares for parameter estimation. A new autoregressive model, incorporating a flexible form for latent variables used to depict homophily, is proposed. A first-order differencing method is developed for the estimation of influence, requiring only the influence part of the model to be correctly specified. When the homophily part is correctly specified, admitting a semiparametric form, we leverage and generalize the recent notion of neighbor smoothing for parameter estimation, bypassing the need to specify the generative mechanism of the network. A new theory is developed to show that all the estimated parameters are consistent and asymptotically normal. The efficacy of the approach is confirmed via extensive simulations and an analysis of a social media dataset.
