On a multiple observations model in discriminant analysis

Abstract

In this paper the multiple observations, or two factor mixed hierarchical, model for the classification problem has been studied. Under this model the Bayes classification statistic and some of its properties ;re discussed. it will be xeil that the muliipie observations modei includes the fixed eifects modei as a special case and bears an interesting relationship to the random effects model, The mul!iplr observations model has a potential application in a variety of fields. In medicine, for example, a patient to be classified as "healthy" versus "non-healthy" might be observed on several attributes at multiple points in time. If individuals from the two populations differ in the variability of these measurements over time, this model can exploit that difference as an aid to classification. Multivariate normality is assumed throughout this paper.

Decision theory classification of high-dimensional vectors based on small samples

“…To describe these relations we use the matrix-variate normal distribution (Gupta and Nagar 2000). To the best of our knowledge, matrix-variate normal distributions have been used in classification only in the context of repeated measurements, where the same set of n variables is observed on each sample member more than once (Choi 1972;Gupta 1986;Gupta and Logan 1990). However, in the majority of practical cases each sample is observed only once, and this is the case that we study in the present paper.…”

A new classification rule for incomplete doubly multivariate data using mixed effects model with performance comparisons on the imputed data

“…This step draws  (t+1) from P(Â|y obs ; y (t+1) mis ). This creates a Markov chain (y (1) mis ;  (1) ); (y (2) mis ;  (2) ); (y (3) mis ;  (3) ); : : :…”

Classification Based on Multivariate Repeated Measures with Time Effect on Mean Vector and an AR (1) Correlation Structure on the Repeated Measures

“…Gupta (1986) extended the method in the multivariate scenario. Gupta and Logan (1990) have also studied the classical approach to this problem. Also, Bayesiau predictive discrimination using a diffuse prior for the mean vector and the variance-covariance matrices has been studied by Logan and Gupta (1993).…”