Alimatu
Agongo

Machine learning in promotion time cure models

I investigate machine learning methods for the incidence component of promotion time cure models. This work compares regression, splines, support vector machines, decision trees, random forests, neural networks, and gradient boosting.

Evaluation focuses on cure-probability estimation, survival estimation, and predictive discrimination through simulation studies and real-data analysis.

Cure models for interval-censored data

I study EM-based estimation for promotion time cure models when event times are observed within intervals. Ongoing extensions address mixed-case censoring, including left-censored, interval-censored, right-censored, and exact event-time observations.

I am also exploring parametric and semiparametric approaches to the latency distribution.

Machine learning for exchange-rate forecasting

At KNUST, I compared support vector machine and long short-term memory models for forecasting Ghana’s exchange rate.