Suggestion for link prediction using TERGM

[eugeneangty]

Hi Pavel and David,

This is Eugene. I spoke to both of you at INSNA conference about my idea of using different weights on past observations and the derived coefficient estimates for link prediction. So, I'm opening up this suggestion here. I'm also thinking if there is a neater way to compute the estimates of any interval between a given observation and its immediate predecessor, perhaps for a difference-in-difference approach, to check for the relative change of a statistic across time.