AfricArXiv
Not a member yet
1453 research outputs found
Sort by
Support Vector Classifiers in scikit-learn: Mathematical Detail, Part II
No full textWe present mathematical detail pertaining to the theory of soft-margin support vector classifiers, designated C-SVC, as used in scikit-learn. We discuss the character of C-SVC, particularly with regard to the penalty term. We construct the primal problem and, thereafter, derive the dual problem. We introduce the notion of nonlinear classifiers and describe the so-called kernel trick. Additionally, we show how the primal problem can be derived from the dual problem. The paper is the second in a series and is intended to be educational in nature
Democratising the use of video in African Higher Education Institutions: The Africa Digital Campus use case.
No full text
New wine in new wineskins: Research Data Management (RDM) by Academic Libraries as innovative value added service in higher education communities
No full text
AfricArXiv - the pan-African Open Access Portal
No full textThis presentation was delivered at Open Repositories 2023 Conference as part of Panel 5: Supporting Trusted OA Usage Data Reporting and Analytics through Shared Infrastructure
Impact of Microbial Load on the Bioremediation of Crude Oil Polluted Soil
Get PDFThe effect of microbial load and nutrient (NPK 15:15:15) on crude oil polluted soil was investigated in this study. 10% crude oil polluted soil for the samples were impacted with Aspergillus niger of concentrations 1×106 cfu/g, 2×106 cfu/g, 5×106 cfu/g, 1×107 cfu/g, 2×107 cfu/g respectively and NPK 15:15:15 were introduced to five samples which served as test samples, and the other as control sample. Parameters such as pH, Residual Hydrocarbon Content (RHC), Total Microbial Count (TMC) were monitored over a period of seven (7) weeks. The results obtained show an increase in pH for the samples. The Total Hydrocarbon Content (THC) decreased for all the samples. The control sample dropped from 183-27mg/kg (85% drop), sample impacted with 1×106 cfu/g dropped from 179-10mg/kg (94% drop), sample impacted with 2×106 cfu/g dropped from 181-9mg/kg (95% drop), sample impacted with 5×106 cfu/g dropped from 181-8mg/kg (96% drop), sample impacted with 1×107 cfu/g dropped from 180-6mg/kg (97% drop) and the sample impacted with 2×107 cfu/g dropped from 184-4mg/kg (98% drop). From the results, bioremediation strategy was found to be very effective
Examining the bilingual phonological processing and reader profiles for (non)readers of three consistent orthographies: evidence from South Africa
No full textBackground: Developmental delays in phonological awareness (PA) and rapid autotomized naming (RAN) contribute to reading difficulties, even in consistent orthographies. In South Africa, evidence suggests that these skills correlate with reading in African languages. However, the extent to which PA and RAN contribute to different reading profiles is not known and little information exists about whether reading profiles are similar across South African languages.
Objectives: This longitudinal study examined the phonological processing and reader profiles of isiXhosa, isiZulu and Northern Sotho speaking children at two points in time to determine whether consistent profiles emerged.
Method: 59 isiXhosa- and 58 isiZulu-speaking children (analysed as one Nguni group) and 131 Northern-Sotho speaking children were assessed at different points in the foundation phase. Children completed a battery of tests assessing PA, RAN, word reading accuracy, and text reading fluency in their first language and English.
Results: Latent profile analysis revealed four profiles found in each group and time point: reader, developing reader, delayed PA and double deficit. These groups differed clearly in terms of PA and RAN.
Conclusion: The findings support the universality of PA and RAN for (bilingual) reading and support the need for explicit and systematic phonics instruction for reading development
AfricArXiv, Diamond Open Access, Sustainability & Funding
No full textPresentation held at the II DIAMOND OPEN ACCESS CONFERENCE, October 25 - 26, 2023. Toluca, Mexico
Plenary session: Sustainability and funding of diamond open access
Capacity building and sustainable finances will be important for the future development of diamond open access. This session discusses how we can further strengthen, sustain, and fund a diamond open access scholarly communication model
La segmentation non linguistique permet-elle de prédire l'alphabétisation dans un enseignement en seconde langue (L2) ? L'apprentissage statistique dans les écoles primaires ivoiriennes
No full textSupplemental Materials: https://osf.io/udzm7/L’Apprentissage Statistique (SL) est un mécanisme d’apprentissage qui ne dépend pas directement de la connaissance d’une langue, mais prédit des résultats de langage et d’alphabétisation pour enfants et adultes. Les recherches liant l’apprentissage statistique et alphabétisation n’ont pas concerné les enfants qui apprennent pour la première fois à lire dans leur seconde langue (L2), commune dans les écoles primaires à travers le monde entier. Plusieurs études ont établi un lien entre la seconde langue et l’alphabétisation des enfants en Australie, en Chine, en Europe, et les États Unis., et nous avons préenregistré une adaptation pour la Côte d’Ivoire, où des enfants sont scolarisés en français et parlent une langue locale à la maison. Après avoir recruté 117 élèves en classe de CM2 des écoles primaires de plusieurs villages, nous avons fait le test pour des corrélations supérieures à 0,3 entre la langue maternelle et alphabétisation avec un taux allant de 80 et 90%. Nous n’avons trouvé aucune preuve pour ces corrélations, mais l’apprentissage statistique visuel (VSL) a été corrélé avec la conscience phonologique en seconde langue (L2). Bien que ce résultat puisse suggérer un rôle de la seconde langue dans les compétences émergentes en seconde langue (L2), il souligne la nécessité d'inclure les contextes d'acquisition de la seconde langue (L2) dans la recherche sur l'alphabétisation. Cet article est une traduction de : "Does non-linguistic segmentation still predict literacy in an L2 education? Statistical learning in Ivorian primary schools" https://osf.io/preprints/africarxiv/fg8s9