The present paper announces and demonstrates several properties of thcorrelation matrix implied by a recursive path model. The main result is that this matrix is affine with respect to the model parameters. These properties concern especially the estimation step in which one should seeks for numerical value for each parameter. Several minimization algorithms are available in softwares for this task, which use iterative procedure that generally requiresthe computation of successive gradient vectors and the Hessian matrices. The properties established in this paper provide more essence and exibility within these types of computation.
Properties of the correlation matrix implied by a recursive path model obtained using the Finite Iterative Method / Hadri, Z. E.; Iaousse, M.; Hanafi, M.; Dolce, P.; Kettani, Y. E.. - In: ELECTRONIC JOURNAL OF APPLIED STATISTICAL ANALYSIS. - ISSN 2070-5948. - 13:2(2020), pp. 413-435. [10.1285/i20705948v13n2p413]
Properties of the correlation matrix implied by a recursive path model obtained using the Finite Iterative Method
Dolce P.;
2020
Abstract
The present paper announces and demonstrates several properties of thcorrelation matrix implied by a recursive path model. The main result is that this matrix is affine with respect to the model parameters. These properties concern especially the estimation step in which one should seeks for numerical value for each parameter. Several minimization algorithms are available in softwares for this task, which use iterative procedure that generally requiresthe computation of successive gradient vectors and the Hessian matrices. The properties established in this paper provide more essence and exibility within these types of computation.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
