TY - GEN
T1 - Multivariate regression model selection with KIC for extrapolation cases
AU - Seghouane, Abd Krim
PY - 2005
Y1 - 2005
N2 - The Kullback Information Criterion, KIC and its multivariate bias-corrected version, KICVC are two alternatively developed criteria for model selection. The two criteria can be viewed as estimators of the expected Kullback symmetric divergence. In this paper, a new criterion is proposed in order to select a well fitted model for an extrapolation case. The proposed criterion is named, PKIC, where "P" stands for prediction, and is derived as an exact unbiased estimator of an adapted cost function that is based on the Kullback symmetric divergence and the future design matrix. PKIC is an unbiased estimator of its cost function assuming that the true model is correctly specified or overfilled. A simulation study illustrating that model selection with PKIC performs well for some extrapolation cases is presented.
AB - The Kullback Information Criterion, KIC and its multivariate bias-corrected version, KICVC are two alternatively developed criteria for model selection. The two criteria can be viewed as estimators of the expected Kullback symmetric divergence. In this paper, a new criterion is proposed in order to select a well fitted model for an extrapolation case. The proposed criterion is named, PKIC, where "P" stands for prediction, and is derived as an exact unbiased estimator of an adapted cost function that is based on the Kullback symmetric divergence and the future design matrix. PKIC is an unbiased estimator of its cost function assuming that the true model is correctly specified or overfilled. A simulation study illustrating that model selection with PKIC performs well for some extrapolation cases is presented.
UR - http://www.scopus.com/inward/record.url?scp=33745955575&partnerID=8YFLogxK
U2 - 10.1109/IJCNN.2005.1556040
DO - 10.1109/IJCNN.2005.1556040
M3 - Conference contribution
SN - 0780390482
SN - 9780780390485
T3 - Proceedings of the International Joint Conference on Neural Networks
SP - 1292
EP - 1295
BT - Proceedings of the International Joint Conference on Neural Networks, IJCNN 2005
T2 - International Joint Conference on Neural Networks, IJCNN 2005
Y2 - 31 July 2005 through 4 August 2005
ER -