Acanthaceae are among the most taxonomically diverse, geographically widespread, and morphologically and ecologically variable lineages of flowering plants. Most modern workers have estimated more than 4000 species and potentially more than 5000 species worldwide, thus placing Acanthaceae among the 12 or so most diverse families of angiosperms. This diversity is marked by exceptional morphological variation, particularly with respect to floral forms, growth forms, and pollen types. The present work represents a synthesis of knowledge generated over the past two decades on the taxonomy and systematics of this complex plant family. We place all 191 accepted genera within a revised classification of the family. Dichotomous keys (nine in total) to recognize the major lineages of Acanthaceae are presented together with geographically partitioned keys to all genera, covering (a) Africa, Madagascar, the Mediterranean region, and Arabia; (b) Asia and Australasia; and (c) the Americas. Finally, we validate several new tribes, subtribes, and genera, and provide new combinations for species where generic delimitation has changed. Our hope is that the present contribution serves to benefit future research on the systematics of Acanthaceae and provides a foundation upon which future classification efforts can be built.

A taxonomic account of Acanthaceae subfamily Nelsonioideae based on morphological and phylogenetic data treats five genera with 172 species: Anisosepalum (3), Elytraria (21), Nelsonia (2), Saintpauliopsis (1), and Staurogyne (145). Two other currently recognized genera, Gynocraterium and Ophiorrhiziphyllon, are included within Staurogyne, and the new combinations, Staurogyne guianensis and S. macrobotrya, are proposed. Probable apomorphic and other diagnostic macro- and micromorphological characters are discussed relative to the subfamily and genera. Characters of the inflorescence, androecium (especially pollen), and seed show important phylogenetic and diagnostic signal. A key to genera, generic descriptions and discussions, illustrations, and distribution maps are provided. Lists of currently recognized species for each genus include synonymies and distributions by country.

Classification algorithms have recently found applications in computational physics for the selection of numerical methods or models adapted to the environment and the state of the physical system. For such classification tasks, labeled training data come from numerical simulations and generally correspond to physical fields discretized on a mesh. Three challenging difficulties arise: the lack of training data, their high dimensionality, and the non-applicability of common data augmentation techniques to physics data. This article introduces two algorithms to address these issues: one for dimensionality reduction via feature selection, and one for data augmentation. These algorithms are combined with a wide variety of classifiers for their evaluation. When combined with a stacking ensemble made of six multilayer perceptrons and a ridge logistic regression, they enable reaching an accuracy of $90\%$ on our classification problem for nonlinear structural mechanics.

Four species of Anisotes Nees (Acanthaceae) are described from northern and west-central Madagascar: A. hygroscopicus T. F. Daniel, Letsara & Martín-Bravo, A. perplexus T. F. Daniel, Letsara & Martín-Bravo, A. subcoriaceus T. F. Daniel, Letsara & Martín-Bravo, and A. venosus T. F. Daniel, Letsara & Martín-Bravo. A key to the six species of the genus known from Madagascar, all of them endemic to the island, is provided. Morphological features previously unknown in the genus are noted for A. hygroscopicus and A. venosus (hygroscopic trichomes on seeds), A. subcoriaceus (2-colporate, pseudocolpate pollen lacking insulae), and A. perplexus (2-pororate pollen). None of these species can be treated with certainty in any of the currently recognized sections of Anisotes. Data pertinent to the conservation status of each species are provided.