RECURSIVE ESTIMATION AND RESIDUALS

If observations of a regression relationship have a natural ordering, the estimates of the regression parameters based on the first t observ ations are called recursive estimates. The paper discusses how to obta in these estimates beginning at t = 0, and the relationship of the cha nges of the estimates to recursive residuals, which are proportional t o the differences between the observation of the response variable at time t and its prediction using information up to time t - 1. An exten sion of these ideas to generalized linear models is suggested.