Coordinators: Patrick Weke, University of Nairobi ([email protected]), John Mango, Makerere University ([email protected]),
Researchers involved: Elena Gaura ( Coventry University), Denis Ndanguza ( University of Rwanda), Niklas Wahlström (Uppsala University), Matthew England (Coventry University); Jean Barbier(ICT)
As AI moves ahead, there is a pressing need to gain a firmer understanding of the mathematical methods/tools used in the development of AI. These methods help in the evaluation of problems, application of correct and reliable tools of analysis, and correct interpretation of results. Some of the topics to be considered are network analysis, graph theory, categorical theory, optimisation theory, PDEs, stochastic processes, physics informed neural networks, control theory, and quantum inspired methods.
We suggested that (i) training in the mathematical methods from foundation, intermediate and advanced levels be incorporated; (ii) those mathematical methods be included as core-competencies for researchers and those trainees in the doctoral programmes; and (iii) develop modules covering the methods and use them as benchmarks for curriculum development by participating universities.