Heckman, N. E., and Ramsay, J. O. (2000). Penalized regression with model-based penalties, Can. J. Statist., 28, 241--258. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Improving Penalized Least Squares Through Adaptive Selection Of.. - Beran (2001) (Correct)
....estimators have relatively low estimated risk among this group of competing shrinkage adaptive estimators because the underlying trend in the melanoma data is roughly linear. The plotted shrinkage adaptive estimators capture ripples in melanoma incidence that are associated with the sunspot cycle. Heckman and Ramsay (2000) obtained similar fits to this data with continuous spline penalized least squares, using di#erential penalty operators analogous to D d and a non automatic choice of penalty weight. Their treatment also considered a penalty di#erential operator that annihilates sinusoids of specified frequency. ....
Heckman, N. E., and Ramsay, J. O. (2000). Penalized regression with model-based penalties, Can. J. Statist., 28, 241--258.
Nonparametric Smoothing Using State Space Techniques - Brown, de Jong (Correct)
....or nonstationarity by di#erencing to stationarity. Di#useness and di#erencing to stationarity are discussed in Sections 2 and 5. The Reinsch algorithm computes smooths via quantities here called smoothations. The general properties and uses of smoothations are explored in Section 4. Heckman Ramsay (2000) express the favoured aspects of a model and fit using PLS constructs. By using the connection between PLS and state space smoothing, these constructs can be translated to statements about the underlying state space model. Di#erent aspects of a PLS penalty matrix are made explicit by considering ....
....The sequential covariance structure induced by the model thus largely dictates the overall fit. Figure 5: Parametric and non parametric fit components for dental smooth. 500 1000 1500 2000 2500 3000 3500 0 200 400 600 800 1000 force The parametric component is closely related to what Heckman Ramsay (2000) define as the favoured model in relation to the PLS problem (1) The favoured model for s is the set of s as y varies and # ##.As#in (1) increases, smoothness becomes the overriding concern and s has zero penalty and Ds =0. Thusthe favoured model is the null space of D or, equivalently, ....
N. E. Heckman & J. O. Ramsay (2000). Penalized regression with model-based penalties. The Canadian Journal of Statistics, 28, in press.
Differential Equation Models for Statistical Functions - Ramsay (2000) (1 citation) Self-citation (Ramsay) (Correct)
....p = exp C 1 D 1 exp D 1 w C 0 expC 1 D 1 exp D 1 w . 3. DIFE S AS ROUGHNESS PENALTIES We have already observed in Section 2.1 that a DIFE can define a low dimensional fit to data that lies at one end of a spectrum of data smooths and that have at the other extreme a saturated model. Heckman Ramsay (2000) consider smoothing discrete data Y i ,i=1, N using the criterion F # (Y g) N # i Y i g(x i ) 2 # # Lg(x) 2 dx. Here, L is a linear di#erential operator of order m Lg(x) # 0 (x)g(x) # 1 (x)Dg(x) # m 1(x )D m 1 g(x) D m g(x) 13) The coe#cient functions # j (x) j ....
N. E. Heckman & J. O. Ramsay (2000). Penalized regression with model-based penalties. The Canadian Journal of Statistics, 28, in press.
CriSP - a Tool for Bump Hunting - Harezlak, Heckman (1999) Self-citation (Heckman) (Correct)
No context found.
Heckman, N.E. and Ramsay, J.O. (in press), "Penalized Regression with Model-Based Penalties," Canadian Journal of Statistics. Minnotte, M.C. (1997), "Nonparametric Testing of the Existence of Modes," The Annals of Statistics, 25, 1646-1660.
The Theory and Application of Penalized Least Squares Methods.. - Nancy Heckman (1997) (2 citations) Self-citation (Heckman Ramsay) (Correct)
....on Fredholm integral equations of the first kind, that is, E(Y j ) g(t j ) where g(t j ) R b a H(s; t j ) s) ds j F j ( with H known. Such data can arise in tomography. For other applications, see the references in Wahba. Wahba also takes L = 00 . Ansley, Kohn, and Wong (1993) and Heckman and Ramsay (1996) demonstrate the usefulness of using L s other than L = 00 . Figure 1, taken from the Heckman and Ramsay paper, shows two estimates of a regression function for the incidence of melanoma in males. The data, described in Andrews . ....
Heckman, N. and Ramsay, J.O. (1996) Penalized Regression with ModelBased Penalties.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC