A Large Dimensional Study of Regularized Discriminant Analysis Classifiers
This work provides theoretical insights into classifier performance for high-dimensional data, with practical applications in optimizing regularization parameters, though it is incremental as it builds on existing random matrix theory.
The authors conducted a large dimensional analysis of regularized discriminant analysis classifiers under Gaussian mixture assumptions, showing that the asymptotic classification error approaches a deterministic quantity dependent on class means, covariances, and dimensions, which accurately predicts performance on real USPS data sets.
This article carries out a large dimensional analysis of standard regularized discriminant analysis classifiers designed on the assumption that data arise from a Gaussian mixture model with different means and covariances. The analysis relies on fundamental results from random matrix theory (RMT) when both the number of features and the cardinality of the training data within each class grow large at the same pace. Under mild assumptions, we show that the asymptotic classification error approaches a deterministic quantity that depends only on the means and covariances associated with each class as well as the problem dimensions. Such a result permits a better understanding of the performance of regularized discriminant analsysis, in practical large but finite dimensions, and can be used to determine and pre-estimate the optimal regularization parameter that minimizes the misclassification error probability. Despite being theoretically valid only for Gaussian data, our findings are shown to yield a high accuracy in predicting the performances achieved with real data sets drawn from the popular USPS data base, thereby making an interesting connection between theory and practice.