Selected Projects
This page highlights selected research and applied projects spanning representation learning, statistical genetics, survival analysis, machine learning, optimization, and mathematical sciences.
Current and Recent Research Projects
Representation Learning for Unstructured Genetic Data
University of CalgaryPhD Research Project
This research develops novel biostatistical and representation-learning methods for identifying genetic factors underlying complex diseases, including cancer and cardiovascular disorders. The project bridges statistical genetics, deep learning, and explainable AI to improve the interpretation of disease biology, polygenic architecture, and heritability.
Analysis of Multi-Scale Data
University of CalgaryResearch Project
This project focused on the analysis and interpretation of multi-scale data using a real-world survival-data application. The goal was to understand how information measured at different biological or clinical scales can be integrated into a coherent statistical analysis framework.
Survival Analysis of Heart Failure Patients
University of CalgaryResearch Project
This project applied survival-analysis methods to heart-failure patient data in order to study time-to-event outcomes and identify clinical factors associated with patient prognosis.
Predicting Phenotypes from Genotypes
University of CalgaryResearch Project
This project compared multiple statistical and machine-learning models for predicting phenotypic traits from genotype data. Methods included regularized regression approaches such as ridge regression and lasso, with the broader goal of evaluating how genetic variation can be used for predictive modeling of observable traits.
Graduate and Training Projects
Optimization and Generalization of Shallow Neural Networks with Quadratic Activation Functions
AIMS Ghana / AMMIM.Sc. in Machine Intelligence Thesis Project
This project was completed as part of the African Master’s in Machine Intelligence program in Ghana. It investigated optimization and generalization properties of shallow neural networks using quadratic activation functions.
Completed under the AMMI program, supported by Google and Facebook.
Accelerating Stochastic Gradient Descent using Predictive Variance Reduction
Gene Golub SIAM Summer SchoolSummer School Research Project
This project was completed as part of the South Africa group at the Gene Golub SIAM Summer School on the theory and practice of deep learning. The work explored predictive variance reduction methods for accelerating stochastic gradient descent.
Abelian Extensions and Crossed Modules for Lie Algebras
AIMS SenegalM.Sc. in Mathematical Sciences Thesis Project
This project was completed for the M.Sc. in Mathematical Sciences at the African Institute for Mathematical Sciences in Senegal. It studied abelian extensions and crossed modules in the context of Lie algebras.
(Co)homologie des espaces de configuration
University of DschangMaster’s Degree in Mathematics Project
This project was completed for the Master’s degree in Mathematics at the University of Dschang, Cameroon. It focused on the cohomology and homology of configuration spaces.