Mathematical Model of Disease Progression and Application of Machine Learning Algorithms for Predicting Disease Stages

Abstract

Accurately diagnosing a patient's disease stage is critical to deal with the varying symptoms and applying appropriate therapeutics at different stages.  Therefore, we used and compared machine learning algorithms on synthetic data and classified disease stages.  This study was performed on the Google Colab environment with Python's machine learning libraries (Sklearn).  The simulated data resembled one that might be reasonably seen in real patients' data since it contained both temporal variation and various noise levels.  Three types of machine learning algorithms were used: Nearest Neighbor, Decision Tree, and Neural Network.  These algorithms could classify whether the current stage of disease progression was at its early, mid, or late stages under different noise levels and other time intervals.  The neural network algorithm showed the best performance.  Although this study used a synthetic data set, it demonstrated how machine learning could be applied in the medical field to enhance patient care.