Clinical ML · B.Sc. graduation project · Team project · publication under revision
Breast Cancer Metastasis Prediction
I contributed the ML modelling, feedback-tracking web application, and deployment for a clinical metastasis-prediction system.
01
The challenge
Work with a clinically structured dataset containing substantial categorical information while keeping predictions, feedback, and later model retraining connected in one system.
02
The approach
I developed the CatBoost modelling work and a Django/PostgreSQL application for prediction access, visualization, feedback tracking, and retraining support. The wider graduation project was completed by a multi-person team.
A model selected for the data
CatBoost was chosen in part because much of the clinical dataset was categorical, rather than because it was simply the most fashionable option.
My contribution boundary
My individually confirmed work covers the ML modelling, the feedback-tracking web application, and its deployment within a larger team graduation project.
03
The outcome
A complete clinical ML and web workflow that is now in the publication phase and under revision; I do not present the wider team project as solo work.