Skip to main content

Prudence Djagba, Ph.D.

Prudence Djagba
  • Assistant Professor
  • LB Course Subject Area: Mathematics
  • Department of Finance
  • Pronouns: he/him

BIOGRAPHY

I am a mathematician interested in advances in academic and practical fields afforded by machine learning and artificial intelligence. My primary research area is in nearfields and near-vector spaces theory, which are branches of non-distributive and non-linear algebra. I also conduct research on the applications of machine learning models and statistical models in quantitative finance and science education. I teach the course FI 491-005, "AI in Finance".

EDUCATION

Ph.D., Mathematics from Stellenbosch University, Stellenbosch, South Africa 

M.Sc., Mathematics from Stellenbosch University, Stellenbosch, South Africa

RESEARCH

Currently my work includes:

  • Exploring the fine tuning of Deep Learning models and Data Augmentation to accurately analyze student descriptions of scientific models, compared to human experts, in analytic scoring predictions based on text classification in NGSS Classroom.
  • Combination of Graph Neural Network with Probabilistic Graphical Models to predict experimental quantities of molecules in Computational Chemistry.
  • Exploring Large Language Models for Financial Applications: Techniques, Performance, and Challenges
  • Classifying Diophantine Triples through Machine Learning Algorithms

SELECTED PUBLICATIONS

  • Djagba, K-T Howell "The subspace of the finite-dimensional Beidleman near-vector spaces", Journal of Linear and Multilinear Algebra, Volume 68, Pages 2316-2336, 2020, doi.org/10.1080/03081087.2019.1582610.
  • P. Djagba "On the generalized distributive set of a finite nearfield", Journal of Algebra, Volume 542, Pages 130-161, 2020, doi.org/10.1016/j.jalgebra.2019.09.020
  • P. Djagba, A. Prins "On Linear Maps and Seed Sets of Beidleman Near-Vector Spaces", Palestine Journal of Mathematics, Vol 14(1)(2025), 195–212, arXiv:2310.05948.
  • L. Kaldaras, T. Li, P. Djagba, K. Haudek, J. Krajcik "Learning Progression-Guided AI Evaluation of Scientific Models To Support Diverse Multi-Modal Understanding in NGSS Classroom ", Accepted and published at NARST Conference (March 23-26, 2025) Washington DC, arxiv:2509.18157
  • P. Djagba, A. Zeleke and V. Rakotonarivo "Hybrid Deep Learning Models for the Prediction of Some Experimental Quantities of Molecules ", accepted and presented at ICMLA (December 3-5, 2025) Florida.