OKAPI: Automated Deep Ensemble Design Through Genetic Programming
A genetic programming algorithm that evolves tree-structured fusion schemas over frozen base-model predictions, automating model selection and minimizing ensemble size.
Damian Kucharski
ML Researcher · ML Systems · Medical & Space Data
I’m an ML Researcher working on machine learning systems and applied research across medical and space data. I also apply generative AI and LLM-powered tooling to real engineering workflows. My background also covers multivariable optimization and constrained optimization problems. I have co-authored peer-reviewed publications in machine learning and medical imaging and am a co-inventor on patents for automated medical image analysis.
Recent Publications
Detecting cardiovascular diseases using ECG scans and explainable artificial intelligence
An explainable AI framework for quantifying the stability of deep learning models analyzing ECG scans under controlled image perturbations.
Interpretable arrhythmia detection in ECG scans using deep learning ensembles: a genetic programming approach
A deep learning ensemble framework for detecting cardiac arrhythmias and predicting atrial fibrillation recurrence from ECG scans.