We develop a dynamic Bayesian beta model for modeling and forecasting single time series of proportions. This work is related to the class of the so called dynamic generalized linear models (DGLM). We use non-conjugate priors and some forms of approximate Bayesian analysis, including Linear Bayesian estimation. Some applications to both real and simulated data are provided.

We present a novel multilabel/ranking algorithm working in partial information settings. The algorithm is based on 2nd-order descent methods, and relies on upper-confidence bounds to trade-off exploration and exploitation. We analyze this algorithm in a partial adversarial setting, where covariates can be adversarial, but multilabel probabilities are ruled by (generalized) linear models. We sho...

We study robust high-dimensional estimation of generalized linear models (GLMs); where a small number k of the n observations can be arbitrarily corrupted, and where the true parameter is high dimensional in the “p n” regime, but only has a small number s of non-zero entries. There has been some recent work connecting robustness and sparsity, in the context of linear regression with corrupted o...

We propose a Generalized Dantzig Selector (GDS) for linear models, in which any norm encoding the parameter structure can be leveraged for estimation. We investigate both computational and statistical aspects of the GDS. Based on conjugate proximal operator, a flexible inexact ADMM framework is designed for solving GDS. Thereafter, non-asymptotic high-probability bounds are established on the e...

In the last four decades, human fertility has declined across the globe, a trend that policy experts predict will continue. As a result, international migration will increasingly affect the demographic future of many nations. Joel Cohen et al. devised a generalized linear model to predict the yearly number of regional and international migrants. The model includes projected demographic variable...

Nonparametric methods for the estimation of the link function in generalized linear models are able to avoid bias in the regression parameters. But for the estimation of the link typically the full model, which includes all predictors, has been used. When the number of predictors is large these methods fail since the full model can not be estimated. In the present article a boosting type method...

An essential step in understanding the processes underlying the general mechanism of perceptual categorization is to identify which portions of a physical stimulation modulate the responses of our perceptual system. More specifically, in the context of speech comprehension, it is still unclear what information is used to categorize a speech stimulus as one phoneme or another. Here we propose to...

This paper aims to identify different potential factors associated with childbearing pattern among the ever-married women in Bangladesh. Childbearing pattern is directly related to fertility level and rapid population growth is the major consequence of more childbearing. Bearing more children affects adversely on social and economic opportunities and produces substantial risks to the health of ...