The problem of using time-varying trajectory data measured on many process variables over the finite duration of a batch process is considered. Multiway principal-component analysis is used to compress the information contained in the data trajectories into low-dimensional spaces that describe the operation of past batches. This approach facilitates the analysis of operational and quality-control problems in past batches and allows for the development of multivariate statistical process control charts for on-line monitoring of the progress of new batches. Control limits for the proposed charts are developed using information from the historical reference distribution of past successful batches. The method is applied to data collected from an industrial batch polymerization reactor.

https://www.jstor.org/stable/pdfplus/1269152.pdf

Multivariate SPC charts for monitoring batch processes

Anomalies are unusual and significant changes in a network's traffic levels, which can often span multiple links. Diagnosing anomalies is critical for both network operators and end users. It is a difficult problem because one must extract and interpret anomalous patterns from large amounts of high-dimensional, noisy data.In this paper we propose a general method to diagnose anomalies. This method is based on a separation of the high-dimensional space occupied by a set of network traffic measurements into disjoint subspaces corresponding to normal and anomalous network conditions. We show that this separation can be performed effectively by Principal Component Analysis.Using only simple traffic measurements from links, we study volume anomalies and show that the method can: (1) accurately detect when a volume anomaly is occurring; (2) correctly identify the underlying origin-destination (OD) flow which is the source of the anomaly; and (3) accurately estimate the amount of traffic involved in the anomalous OD flow.We evaluate the method's ability to diagnose (i.e., detect, identify, and quantify) both existing and synthetically injected volume anomalies in real traffic from two backbone networks. Our method consistently diagnoses the largest volume anomalies, and does so with a very low false alarm rate.

/pdf/diagnosing-network-wide-traffic-anomalies-2oqvh3kj5a.pdf

Diagnosing network-wide traffic anomalies

This paper is concerned with the treatment of residuals associated with principal component analysis. These residuals are the difference between the original observations and the predictions of them using less than a full set of principal components. Specifically, procedures are proposed for testing the residuals associated with a single observation vector and for an overall test for a group of observations. In this development, it is assumed that the underlying covariance matrix is known; this is reasonable for many quality control applications where the proposed procedures may be quite useful in detecting outliers in the data. A numerical example is included.

https://www.jstor.org/stable/info/1267757

Control Procedures for Residuals Associated With Principal Component Analysis

We consider the problem of network anomaly detection in large distributed systems. In this setting, Principal Component Analysis (PCA) has been proposed as a method for discovering anomalies by continuously tracking the projection of the data onto a residual subspace. This method was shown to work well empirically in highly aggregated networks, that is, those with a limited number of large nodes and at coarse time scales. This approach, however, has scalability limitations. To overcome these limitations, we develop a PCA-based anomaly detector in which adaptive local data filters send to a coordinator just enough data to enable accurate global detection. Our method is based on a stochastic matrix perturbation analysis that characterizes the tradeoff between the accuracy of anomaly detection and the amount of data communicated over the network.

/pdf/in-network-pca-and-anomaly-detection-135kwzilun.pdf

In-Network PCA and Anomaly Detection

Data analysis is an essential tenet of analytical chemistry, extending the possible information obtained from the measurement of chemical phenomena. Chemometric methods have grown considerably in recent years, but their wide use is hindered because some still consider them too complicated. The purpose of this review is to describe a multivariate chemometric method, principal component regression, in a simple manner from the point of view of an analytical chemist, to demonstrate the need for proper quality-control (QC) measures in multivariate analysis and to advocate the use of residuals as a proper QC method.

/pdf/multivariate-concentration-determination-using-principal-1xbb32us66.pdf

Multivariate concentration determination using principal component regression with residual analysis.

Let Qk = Σk j-1 cj(xj + aj)2 be a definite quadratic form in independent standardized Gaussian variables, xj, EQk = 01. The normalizing transformation (Qk/01)h is investigated, where h is determined by the first three moments of Qk. A new Gaussian approximation to the noncentral chi-square (x2) distribution is found for which the coefficient of skewness is smaller than a cube root transformation in the literature. Our transformation specializes to the cube root transformation of Wilson and Hilferty for the central x2 distribution. The approximation is simple to apply and compares well with other approximations in a number of cases studied numerically.

A Gaussian Approximation to the Distribution of a Definite Quadratic Form

The inequality is shown to hold when the joint distribution of U 1 and U 2 is a member of a class of bivariate Chi-Square distributions given. The relevance of this result to the construction of two-sided simultaneous confidence intervals for variances is noted.

An Inequality for a Class of Bivariate Chi-Square Distributions

Multidimensional Wilson-Hilferty transformations support Gaussian approximations to certain joint distributions of quadratic forms in jointly Gaussian variates. Central and noncentral joint distributions are studied and applications are noted. Parameters of the approximating distributions are given up to terms of order o(v −2) in the degrees of freedom v. Numerical studies validate using these approximations over a range of parameters for an essential subclass of the distributions studied, especially in upper tails.

Approximations to Joint Distributions of Definite Quadratic Forms

Given a sample of n vector observations from a multivariate normal population, Anderson and Roy-Gnanadesikan have given for the variances a set of confidence intervals which are approximate in that a lower bound only is known for the joint confidence coefficient. In the present study, exact procedures are developed in terms of multivariate Chi-Square distributions, and more general approximate procedures are given via Bonferroni's inequality. A numerical investigation suggests that the Bonferroni lower bound is fairly sharp for a variety of parameter values, and it always is superior to the Roy-Gnanadesikan procedure in the bivariate case examined. A lower bound in terms of independent statistics further is examined for a special class of one-sided intervals.

Simultaneous Confidence Intervals for Variances

Efficiencies of Friedman's χ2 r test are studied when observations within blocks are independent and when they are exchangeably dependent. At local alternatives the limiting power is found to increase monotonically with (a) the canonical correlations of distributions having expansions in canonical form, (b) the stochastically decreasing character of the mixing distributions in certain exchangeable sequences, and (c) an index of peakedness in a class of distributions containing exchangeable Cauchy and Gaussian laws.

D. R. Jensen

Papers

A Gaussian Approximation to the Distribution of a Definite Quadratic Form

An Inequality for a Class of Bivariate Chi-Square Distributions

Approximations to Joint Distributions of Definite Quadratic Forms

Simultaneous Confidence Intervals for Variances

Efficiency of Friedman's χ2 r Test Under Dependence