Incremental Linear Discriminant Analysis Using Sufficient Spanning Sets and Its Applications

This paper presents an incremental learning solution for Linear Discriminant Analysis (LDA) and its applications to object recognition problems. We apply the sufficient spanning set approximation in three steps i.e. update for the total scatter matrix, between-class scatter matrix and the projected data matrix, which leads an online solution which closely agrees with the batch solution in accuracy while significantly reducing the computational complexity. The algorithm yields an efficient solution to incremental LDA even when the number of classes as well as the set size is large. The incremental LDA method has been also shown useful for semi-supervised online learning. Label propagation is done by integrating the incremental LDA into an EM framework. The method has been demonstrated in the task of merging large datasets which were collected during MPEG standardization for face image retrieval, face authentication using the BANCA dataset, and object categorisation using the Caltech101 dataset. © 2010 Springer Science+Business Media, LLC.

Vibration and stress analysis in the presence of structural uncertainty

At medium to high frequencies the dynamic response of a built-up engineering system, such as an automobile, can be sensitive to small random manufacturing imperfections. Ideally the statistics of the system response in the presence of these uncertainties should be computed at the design stage, but in practice this is an extremely difficult task. In this paper a brief review of the methods available for the analysis of systems with uncertainty is presented, and attention is then focused on two particular "non- parametric" methods: statistical energy analysis (SEA), and the hybrid method. The main governing equations are presented, and a number of example applications are considered, ranging from academic benchmark studies to industrial design studies. © 2009 IOP Publishing Ltd.

Factor analysis based VTS and JUD noise estimation and compensation

Model based compensation schemes are a powerful approach for noise robust speech recognition. Recently there have been a number of investigations into adaptive training, and estimating the noise models used for model adaptation. This paper examines the use of EM-based schemes for both canonical models and noise estimation, including discriminative adaptive training. One issue that arises when estimating the noise model is a mismatch between the noise estimation approximation and final model compensation scheme. This paper proposes FA-style compensation where this mismatch is eliminated, though at the expense of a sensitivity to the initial noise estimates. EM-based discriminative adaptive training is evaluated on in-car and Aurora4 tasks. FA-style compensation is then evaluated in an incremental mode on the in-car task. © 2011 IEEE.

Evaluating the impacts of iteration on PD processes by transforming task network models into System Dynamics models

Iteration is unavoidable in the design process and should be incorporated when planning and managing projects in order to minimize surprises and reduce schedule distortions. However, planning and managing iteration is challenging because the relationships between its causes and effects are complex. Most approaches which use mathematical models to analyze the impact of iteration on the design process focus on a relatively small number of its causes and effects. Therefore, insights derived from these analytical models may not be robust under a broader consideration of potential influencing factors. In this article, we synthesize an explanatory framework which describes the network of causes and effects of iteration identified from the literature, and introduce an analytic approach which combines a task network modeling approach with System Dynamics simulation. Our approach models the network of causes and effects of iteration alongside the process architecture which is required to analyze the impact of iteration on design process performance. We show how this allows managers to assess the impact of changes to process architecture and to management levers which influence iterative behavior, accounting for the fact that these changes can occur simultaneously and can accumulate in non-linear ways. We also discuss how the insights resulting from this analysis can be visualized for easier consumption by project participants not familiar with simulation methods. Copyright © 2010 by ASME.

Grand Challenges for Data Sensing and Analysis

A diverse group of experts proposed the 9 grand challenges outlined in this booklet. This expert task force was assembled by the ASCE TCCIT Data Sensing and Analysis (DSA) Committee and endorsed by the TRB AFH10(1) Construction IT joint subcommittee at the request of their membership. The task force did not rank the challenges selected, nor did it endorse particular approaches to meeting them. Rather than attempt to include every important goal for data sensing and analysis, the panel chose opportunities that were both achievable and sustainable to help people and the planet thrive. The panel’s conclusions were reviewed by several subject-matter experts. The DSA is offering an opportunity to comment on the challenges by contacting the task force chair via email at becerik@usc.edu.

Factor analysis based VTS discriminative adaptive training

Vector Taylor Series (VTS) model based compensation is a powerful approach for noise robust speech recognition. An important extension to this approach is VTS adaptive training (VAT), which allows canonical models to be estimated on diverse noise-degraded training data. These canonical model can be estimated using EM-based approaches, allowing simple extensions to discriminative VAT (DVAT). However to ensure a diagonal corrupted speech covariance matrix the Jacobian (loading matrix) relating the noise and clean speech is diagonalised. In this work an approach for yielding optimal diagonal loading matrices based on minimising the expected KL-divergence between the diagonal loading matrix and "correct" distributions is proposed. The performance of DVAT using the standard and optimal diagonalisation was evaluated on both in-car collected data and the Aurora4 task. © 2012 IEEE.

Generalized power method for sparse principal component analysis

In this paper we develop a new approach to sparse principal component analysis (sparse PCA). We propose two single-unit and two block optimization formulations of the sparse PCA problem, aimed at extracting a single sparse dominant principal component of a data matrix, or more components at once, respectively. While the initial formulations involve nonconvex functions, and are therefore computationally intractable, we rewrite them into the form of an optimization program involving maximization of a convex function on a compact set. The dimension of the search space is decreased enormously if the data matrix has many more columns (variables) than rows. We then propose and analyze a simple gradient method suited for the task. It appears that our algorithm has best convergence properties in the case when either the objective function or the feasible set are strongly convex, which is the case with our single-unit formulations and can be enforced in the block case. Finally, we demonstrate numerically on a set of random and gene expression test problems that our approach outperforms existing algorithms both in quality of the obtained solution and in computational speed. © 2010 Michel Journée, Yurii Nesterov, Peter Richtárik and Rodolphe Sepulchre.