Committee Members:

Dr. Karl I. Jacob
Dr. Youjiang Wang
Dr. Donggang Yao
Dr. Hamid Garmestani

“Machine Learning Helps to Build Drug Release Kinetic Models”

Abstract:

Long-acting injectables (LAI) are one of the most promising drug delivery systems for the treatment of chronic diseases. Since they can maintain the drug concentration in the target tissue, thus reducing dose frequency and adverse effects as well as improving patient compliance. The use of polymer matrices delivery systems shows an extraordinary diversity in drug development research. But due to the time-consuming experiments and complicated drug release mechanisms, the efficiency of LAI development is strongly restricted. 

This thesis used machine learning to predict the long-period in vitro test profiles based on the datasets collected from published literature. In addition to comparing the accuracy performance of different machine learning algorithms, a combination of empirical mathematic models and machine learning algorithms is further studied in the case to improve the model evaluability.