Several coal-fired power plants (CFPPs) were built in South Africa, mainly in the central Mpumalanga Province, due to an increase in the demand for Eskom, the national power utility, to keep up with socio-economic growth. The CFPPs, of which 90% are owned by Eskom, generate a significant share of the country's electricity but contribute to the air pollution experienced in the country. The paper discusses sulphur dioxide (SO2), nitrogen dioxide (NO2) and particulate matter of size less than 10 micrometre (μm) in diameter (PM10), using data from 2014 to 2018. The statistics revealed higher PM10 concentrations during winter than in summer and spring at the Kriel and Komati sites; associated with the higher contribution of domestic burning. The study's results could influence legislation and policies and help to understand the source of poor ambient air quality by assessing the three pollutants within the area of the selected power plants.

Statistical design of experiments (DoE) aims to develop a near efficient design while minimising the number of experiments required. This is an optimal approach especially when there is a need to investigate multiple variables. DoE is a powerful methodology for a wide range of applications, from the efficient design of manufacturing processes to the accurate evaluation of global optima in numerical studies. The contribution of this paper is to provide a general introduction to statistical design of experiments for a non-expert audience, with the aim of broadening exposure in the applied mechanics community. We focus on response surface methodology (RSM) designs — Taguchi Design, Central Composite Design, Box-Behnken Design and D-optimal Design. These different RSM designs are compared in the context of a case study from the field of polymer composites. The results demonstrate that an exact D-optimal design is generally considered to be a good design when compared to the global D-optimal design. That is, it requires fewer experiments while retaining acceptable efficiency measures for all three response surface models considered. This paper illustrates the benefits of DoE, demonstrates the importance of evaluating different designs, and provides an approach to choose the design best suited for the problem of interest.

Classification requires labelling large sets of data, which is often a time-consuming and expensive process. Active learning is a machine learning technique that has gained popularity in recent years due to its ability to effectively reduce the amount of labelled data required to train accurate models. The success of the active learner heavily relies on the selection of the initial points to initialise the active learning process. In this paper, we compare the performance of the traditional random sampling approach to the maximin Latin Hypercube sampling, conditioned Latin Hypercube sampling, and a modified Latin Hypercube sampling procedure for initialising active learning for the estimation of the logistic regression in binary classification problems. We show that the Latin Hypercube sampling designs outperform random sampling for all the performance measures evaluated. The results are demonstrated using simulated data sets and an actual case study. Specifically, the conditioned Latin hypercube sampling design exhibits high prediction accuracy using a smaller sample size for both heterogeneous and homogeneous classes. In contrast, the modified Latin hypercube sampling design yields the smallest variance of prediction across varying initial sample sizes for both homogeneous and heterogeneous classes. Furthermore, principal component analysis indicates that approximately 10\% of the data is required to develop an accurate and precise logistic regression classifier.