Neuromath is a new and existing field that uses analytical mathematics to explain how artificial intelligence really works. WHile at AIMS, Kenne studied the overparametrisation of the convolutional neural network (CNN) using the concepts of gradient descent (GD) and the neural tangent kernel (NTK). She showed that a CNN learns low frequency before high frequency, which is also called frequency biais.

Thanks to the training that she received at AIMS Cameroon, she currently works as a data scientist at NextView-Gie in CAMEROON.