Two-Stage Orthogonal Least Squares Methods for Neural Network Construction

A number of neural networks can be formulated as the linear-in-the-parameters models. Training such networks can be transformed to a model selection problem where a compact model is selected from all the candidates using subset selection algorithms. Forward selection methods are popular fast subset selection approaches. However, they may only produce suboptimal models and can be trapped into a local minimum. More recently, a two-stage fast recursive algorithm (TSFRA) combining forward selection and backward model refinement has been proposed to improve the compactness and generalization performance of the model. This paper proposes unified two-stage orthogonal least squares methods instead of the fast recursive-based methods. In contrast to the TSFRA, this paper derives a new simplified relationship between the forward and the backward stages to avoid repetitive computations using the inherent orthogonal properties of the least squares methods. Furthermore, a new term exchanging scheme for backward model refinement is introduced to reduce computational demand. Finally, given the error reduction ratio criterion, effective and efficient forward and backward subset selection procedures are proposed. Extensive examples are presented to demonstrate the improved model compactness constructed by the proposed technique in comparison with some popular methods.

Analysis of Extended Partial Least Squares for Monitoring Large-Scale Processes

This paper theoretically analysis the recently proposed "Extended Partial Least Squares" (EPLS) algorithm. After pointing out some conceptual deficiencies, a revised algorithm is introduced that covers the middle ground between Partial Least Squares and Principal Component Analysis. It maximises a covariance criterion between a cause and an effect variable set (partial least squares) and allows a complete reconstruction of the recorded data (principal component analysis). The new and conceptually simpler EPLS algorithm has successfully been applied in detecting and diagnosing various fault conditions, where the original EPLS algorithm did only offer fault detection.

Improved conjugate gradient implementation for least squares support vector machines

As a promising method for pattern recognition and function estimation, least squares support vector machines (LS-SVM) express the training in terms of solving a linear system instead of a quadratic programming problem as for conventional support vector machines (SVM). In this paper, by using the information provided by the equality constraint, we transform the minimization problem with a single equality constraint in LS-SVM into an unconstrained minimization problem, then propose reduced formulations for LS-SVM. By introducing this transformation, the times of using conjugate gradient (CG) method, which is a greatly time-consuming step in obtaining the numerical solution, are reduced to one instead of two as proposed by Suykens et al. (1999). The comparison on computational speed of our method with the CG method proposed by Suykens et al. and the first order and second order SMO methods on several benchmark data sets shows a reduction of training time by up to 44%. (C) 2011 Elsevier B.V. All rights reserved.

Experimentally validated partial least squares model for dynamic line rating

This study presents a model based on partial least squares (PLS) regression for dynamic line rating (DLR). The model has been verified using data from field measurements, lab tests and outdoor experiments. Outdoor experimentation has been conducted both to verify the model predicted DLR and also to provide training data not available from field measurements, mainly heavily loaded conditions. The proposed model, unlike the direct measurement based DLR techniques, enables prediction of line rating for periods ahead of time whenever a reliable weather forecast is available. The PLS approach yields a very simple statistical model that accurately captures the physical performance of the conductor within a given environment without requiring a predetermination of parameters as required by many physical modelling techniques. Accuracy of the PLS model has been tested by predicting the conductor temperature for measurement sets other than those used for training. Being a linear model, it is straightforward to estimate the conductor ampacity for a set of predicted weather parameters. The PLS estimated ampacity has proven its accuracy through an outdoor experiment on a piece of the line conductor in real weather conditions.