We take a fresh look at Theil’s BLUS residuals and ask why they have gone out of fashion. All our simulation experiments indicate that tests based on BLUS residuals have higher power than those based on the more popular recursive residuals, even in those cases (structural breaks) where intuition would favour the recursive residuals. JEL Classification: B23; C10; C13; C20

Tests for change-points for location or regression models are often based on cumulative sums of recursive residuals. In the context of general estimable parameters (functionals of the underlying distributions), such recursive residuals may be defined in terms of recursive U-statistics. Along with some invariance principles for recursive U-statistics, asymptotic properties of some proposed tests...

The use of residuals for detecting departures from the assumptions of the linear model, whether full rank or not, has long been recognized as an important diagnostic tool, once it became computationally feasible to compute them in a straight forward way, and several contributions to the literature have focussed on their underlying properties and their effectiveness. The recursive residuals are ...

Artificial intelligence techniques such as neural networks and fuzzy logic have been widely used in fault detection and diagnosis. Combining these two techniques, referred to as neurofuzzy networks, provides a powerful tool for modelling. B-spline neurofuzzy networks are used to model the residuals. The weights of the networks are trained online using recursive least squares method. Fuzzy rules...

Parameter selection for the criterion weighting matrix is concerned based on the information of both modifying the past estimation residuals and renewing the present estimation residual error . After minimizing the system estimation error, an optimal recursive algorithm is given. In this method the system data record can be used efficiently. The consistency of the new recursive algorithm is ana...