Winsorization for Robust Bayesian Neural Networks

With the advent of big data and the popularity of black-box deep learning methods, it is imperative to address the robustness of neural networks to noise and outliers. We propose the use of Winsorization to recover model performances when the data may have outliers and other aberrant observations. We provide a comparative analysis of several probabilistic artificial intelligence and machine learning techniques for supervised learning case studies. Broadly, Winsorization is a versatile technique for accounting for outliers in data. However, different probabilistic machine learning techniques have different levels of efficiency when used on outlier-prone data, with or without Winsorization. We notice that Gaussian processes are extremely vulnerable to outliers, while deep learning techniques in general are more robust.

Robust Compositional Analysis of Physical Activity and Sedentary Behaviour Data

Although there is an increasing awareness of the suitability of using compositional data methodology in public health research, classical methods of statistical analysis have been primarily used so far. The present study aims to illustrate the potential of robust statistics to model movement behaviour using Czech adolescent data. We investigated: (1) the inter-relationship between various physical activity (PA) intensities, extended to model relationships by age; and (2) the associations between adolescents’ PA and sedentary behavior (SB) structure and obesity. These research questions were addressed using three different types of compositional regression analysis—compositional covariates, compositional response, and regression between compositional parts. Robust counterparts of classical regression methods were used to lessen the influence of possible outliers. We outlined the differences in both classical and robust methods of compositional data analysis. There was a pattern in Czech adolescents’ movement/non-movement behavior—extensive SB was related to higher amounts of light-intensity PA, and vigorous PA ratios formed the main source of potential aberrant observations; aging is associated with more SB and vigorous PA at the expense of light-intensity PA and moderate-intensity PA. The robust counterparts indicated that they might provide more stable estimates in the presence of outlying observations. The findings suggested that replacing time spent in SB with vigorous PA may be a powerful tool against adolescents’ obesity.

Detection of Aberrant Observations in a Background of an Unknown Multidimensional Gaussian Distribution

An exploratory iterative technique for the detection of aberrant observations on a background of a multidimensional Gaussian distribution is described. Its development was motivated by the analysis of a set of three measurements reflecting the acid-base metabolism in the blood of 2,402 intensive care patients. This new, three-dimensional treatment of such data yields a meaningful description. A technical evaluation of the method, using artificially generated data is also presented. It is shown that the model parameters of the underlying Gaussian distributions are determined with good accuracy and that the accuracy with which the contamination is estimated increases with increasing distance of the contaminating observations from the mean.

The direct linear plot. A new graphical procedure for estimating enzyme kinetic parameters

A new plot is described for analysing the results of kinetic experiments in which the Michaelis–Menten equation is obeyed. Observations are plotted as lines in parameter space, instead of points in observation space. With appropriate modifications the plot is applicable to most problems of interest to the enzyme kineticist. It has the following advantages over traditional methods of plotting kinetic results: it is very simple to construct, because it is composed entirely of straight lines and requires no calculation or mathematical tables; the kinetic constants are read off the plot directly, again without calculation; it may be used during the course of an experiment to judge the success of the experiment, and to modify the experimental design; it provides clear and accurate information about the quality of the observations, and identifies aberrant observations; it provides a clear indication of the precision of the kinetic constants; constructed with care, it provides unbiased estimates of the kinetic constants, the same as those provided by a computer program; it may be used to simulate results for illustrative purposes very rapidly and simply.