The PARAFAC-ALS algorithm is the most widely used procedure forapproximating arrays with a trilinear structure because it provides least squaressolutions and delivers consistent outputs. Nonetheless, it is particularly slow atconverging especially under challenging conditions, i.e. data multicollinearity, highfactors’ congruence and over-factoring. This shortcoming can be quite problematicwhen dealing with three-way arrays of large dimensions.More efficient procedures can be employed, such as ATLD, however they are far lessreliable. As an alternative, ATLD and ALS can be combined in a multi-optimizationprocedure in order to increase efficiency without reducing accuracy. This novelapproach has been carried out and tested on artificial and real data.
A PARAFAC-ALS variant for fitting large datasets
GUARINO, MASSIMO
2019-01-01
Abstract
The PARAFAC-ALS algorithm is the most widely used procedure forapproximating arrays with a trilinear structure because it provides least squaressolutions and delivers consistent outputs. Nonetheless, it is particularly slow atconverging especially under challenging conditions, i.e. data multicollinearity, highfactors’ congruence and over-factoring. This shortcoming can be quite problematicwhen dealing with three-way arrays of large dimensions.More efficient procedures can be employed, such as ATLD, however they are far lessreliable. As an alternative, ATLD and ALS can be combined in a multi-optimizationprocedure in order to increase efficiency without reducing accuracy. This novelapproach has been carried out and tested on artificial and real data.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
