Developer
Pneumonia classification
An ensemble pipeline exploring chest X-ray classification, model comparison and evaluation.
Overview
A chest X-ray classification project combining multiple models through an ensemble approach. The project includes preprocessing, augmentation, comparison and confusion-matrix evaluation.
The question
Classification errors have different consequences, so accuracy alone is an incomplete description of a model.
My approach
Compare models using recall, F1 and confusion matrices alongside ensemble predictions.
Outcome and limits
A documented pipeline covering preprocessing, model comparison and recall/F1 evaluation.
An experimental image-classification project, not a clinically validated diagnostic system. Dataset composition and evaluation splits affect how results generalise.
Interested in this work? Get in touch.