Creator |
|
|
Language |
|
Publisher |
|
|
Date |
|
Source Title |
|
Vol |
|
Publication Type |
|
Access Rights |
|
Rights |
|
Related DOI |
|
Related DOI |
|
|
Related URI |
|
|
Related URI |
|
Related HDL |
|
Relation |
|
|
Abstract |
We consider the problem of constructing a functional regression modeling procedure with functional predictors and a functional response. Discretely observed data set are expressed by Gaussian basis ex...pansions for individuals, using smoothing methods. Parameters involved in the functional regression model are estimated by the regularized maximum likelihood method, assuming that coefficient functions are expressed by basis expansions. For the selection of regularization parameters involved in the regularization method, we extend information theoretic and Bayesian model selection criteria for evaluating the estimated model. The proposed modeling strategy is applied to the analysis of real data, predicting functions rather than scalars.show more
|