Challenges of Big Data Analytics in Healthcare


Constantly increasing number of chronic diseases (such as diabetes, cardiovascular diseases, chronic renal failure, cancer etc.) and lack of doctors worldwide, led to the need for paradigm change from delayed interventional to Predictive, Preventive and Personalized Medicine (PPPM). Success stories of the Big Data paradigm and Predictive Analytics in many application areas led to the wide recognition of their high potential impact and benefits (both human and economic ones) in healthcare.  However, there is still a large gap between actual and potential data usage in healthcare, because of numerous challenges: demand for highly accurate and interpretable models, high dimensionality, privacy concerns, the need for collaboration between domain experts and data scientists etc.

In the first part of the talk, I will discuss the challenges and potential benefits of Big Data exploitation in healthcare. In the second part, I will present recent methodological approaches that were applied on real data and published in last several years. These approaches will include sparse predictive modeling, data and knowledge driven feature selection, “white-box” (or “glass box”) algorithm design and meta-learning.

About the lecturer

Milan Vukicevic is an Associate Professor at the University of Belgrade, Faculty of Organizational Sciences. He worked as a Visiting Researcher at the Data Analysis and Biomedical Analytics (DABI) Center at Temple University (2014-2015). His research interests encompass design and development of predictive algorithms and their application in health care predictive modeling. Specific areas of his technical interest include sparse predictive models, fusion of domain knowledge and data-driven methods , parameter and feature selection, predictive analytics on heterogeneous data sources and meta-learning. His work was published in multiple conferences, journals and book chapters. Milan also had several invited talks on bioinformatics and healthcare predictive analytics topics.


Milan Vukićević (SRB)