A space-time wireless communication channels can be decomposed as a set of filters, each consisting of a scalar temporal filter followed by a single spatial signature vector. If only a small number of such filters is necessary to accurately describe the space time channel, we call it a reduced rank channel. We here consider different methods of exploiting this property to improve channel estima...

We propose an approach to multivariate nonparametric regression that generalizes reduced rank regression for linear models. An additive model is estimated for each dimension of a q-dimensional response, with a shared p-dimensional predictor variable. To control the complexity of the model, we employ a functional form of the Ky-Fan or nuclear norm, resulting in a set of function estimates that h...

We propose and justify a new approach to constructing optimal nonlinear transforms of random vectors. We show that the proposed transform improves such characteristics of rank-reduced transforms as compression ratio, accuracy of decompression and reduces required computational work. The proposed transform Tp is presented in the form of a sum with p terms where each term is interpreted as a part...

The multivariate reduced rank regression model plays an important role in econometrics. Examples include co-integration analysis and models with a factor structure. Geweke (1996) provided the foundations for a Bayesian analysis of this model. Unfortunately several of the full conditional posterior distributions, which forms the basis for constructing a Gibbs sampler for the poster distribution,...

In this paper we present a novel, multi-gene approach to time course microarray experiments. One of the advantages of our approach is an explicit modeling of correlation structure among gene expression data. The approach proposed is computationally attractive. We apply the model to the well-known cell-cycle yeast microarray data and present results that compare favorably to the results of the p...