Predicting groundwater availability is important to water sustainability and drought mitigation. Machine-learning tools have the potential to improve groundwater prediction, thus enabling resource planners to: (1) anticipate water quality in unsampled areas or depth zones; (2) design targeted monitoring programs; (3) inform groundwater protection strategies; and (4) evaluate the sustainability of groundwater sources of drinking water. This paper proposes a machine-learning approach to groundwater prediction with the following characteristics: (i) the use of a regression-based approach to predict full groundwater images based on sequences of monthly groundwater maps; (ii) strategic automatic feature selection (both local and global features) using extreme gradient boosting; and (iii) the use of a multiplicity of machine-learning techniques (extreme gradient boosting, multivariate linear regression, random forests, multilayer perceptron and support vector regression). Of these techniques, support vector regression consistently performed best in terms of minimizing root mean square error and mean absolute error. Furthermore, including a global feature obtained from a Gaussian Mixture Model produced models with lower error than the best which could be obtained with local geographical features.

Rooibos is brewed from the medicinal plant Aspalathus linearis. It has a well-established wide spectrum of bio-activity properties, which in part may be attributed to the phenolic antioxidant power. The antioxidant capacity (AOC) of rooibos is related to its total phenolic content (TPC). The relation between TPC and AOC of randomly selected 51 fermented (FR) and 47 unfermented (UFR) rooibos samples was studied after extraction using water and methanol separately. The resulted extracts were assessed using two antioxidant assays, trolox equivalent antioxidant capacity (TEAC) and ferric reducing antioxidant power (FRAP). The results were analyzed using both simple statistical methods and machine learning. The analysis showed different trends of TPC and AOC correlations of FR and UFR samples, depending on the solvent used for extraction. The results of the water extracts showed similar TPC and higher AOC of FR than UFR samples, while the methanolic extracted samples showed higher TPC and AOC of UFR than FR. As a result, the methanolic extracts showed better agreement between TPC and AOC than water extracts. Possible explanations are given for these observed results. Although, the current literature demonstrates direct correlations of the TPC and AOC of rooibos water extracts. This study showed deviation and highlighted the importance of solvent selection and analysis methodology as an important factor in determining the TPC/AOC correlation and subsequently the expectation of the actual health benefits of rooibos herbal tea. In particular, unfermented and fermented samples can be accurately identified on the basis of a combination of assays (any two of TPC, FRAP and TEAC), especially if methanol is the solvent used. Machine learning analysis of assay data provides nearly identical results with classical statistical analytical methods. This is the first report on machine learning analysis and comparison of the TPC and AOC of rooibos herbal tea extracted with methanol and water, and highlights the importance of using methanol as a solvent to evaluate its AOC.

Machine learning (ML) has been utilized to predict climatic parameters, and many successes have been reported in the literature. In this paper, we scrutinize the effectiveness of five widely used ML algorithms in the monthly prediction of seasonal climatic parameters using monthly image data. Specifically, we quantify the predictive performance of these algorithms applied to five climatic parameters using various combinations of features. We compare the predictive accuracy of the resulting trained ML models to that of basic statistical estimators that are computed directly from the training data. Our results show that ML never significantly outperforms the statistical baseline, and underperforms for most feature sets. Unlike previous similar studies, we provide error bars for the relative performance of different predictors based on jackknife estimates applied to differences in predictive error magnitudes. We also show that the practice of shuffling data sequences which was employed in some previous references leads to data leakage, resulting in over-estimated performance. Ultimately, the paper demonstrates the importance of using well-grounded statistical techniques when producing and analyzing the results of ML predictive models.

Visual impairment (VI) affects over two billion people globally, with prevalence increasing due to preventable conditions. To address mobility and navigation challenges, this study presents a multi-platform, multi-sensor Electronic Travel Aid (ETA) integrating a combination of ultrasonic, LiDAR, and vision-based sensing across head-, torso-, and cane-mounted nodes. Grounded in orientation and mobility (OM) principles, the system delivers context-aware haptic and auditory feedback to enhance perception and independence for users with VI. The ETA employs a hardware–software co-design approach guided by proxemic theory, comprising three autonomous components—Glasses, Belt, and Cane nodes—each optimized for a distinct spatial zone while maintaining overlap for redundancy. Embedded ESP32 microcontrollers enable low-latency sensor fusion providing real-time multi-modal user feedback. Static and dynamic experiments using a custom-built motion rig evaluated detection accuracy and feedback latency under repeatable laboratory conditions. Results demonstrate millimetre-level accuracy and sub-30 ms proximity-to-feedback latency across all nodes. The Cane node’s dual LiDAR achieved a coefficient of variation at most 0.04%, while the Belt and Glasses nodes maintained mean detection errors below 1%. The validated tri-modal ETA architecture establishes a scalable, resilient framework for safe, real-time navigation—advancing sensory augmentation for individuals with VI.