Managing and Outsizing Big Data in Health Care Application

• Activity Number: 509
• Type: Contributed
• Date/Time: Wednesday, August 6, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistical Learning and Data Mining
• Author(s): Maoqing Liu*+ and Nasser Fard
• Companies: and Northeastern University
• Keywords: Big Data ; Healthcare ; Spectral Clustering ; Categories

Abstract

Healthcare industries have access to a large volume and variety of data about patients, behaviors, diseases, and treatments. There is a significant need for a data-driven system to discover patterns for better understanding of the impact of human risk behaviors on numerous diseases . By aggregation and integration of data and categorization of research data based on similarities with respect to environment, life style, and health history, the healthcare dataset will provide significant knowledge about the connections of individual behaviors and the risk of developing diseases. This knowledge will help in early prognosis, disease prevention, and disease management. It also provides individual risk profiles, disease plans, and wellness plans based on connections and similarities with other people in the same categories. In this paper, we study spectral clustering algorithm for its application suitability in healthcare data analysis. Many factors, such as scalability, the choice of parameters, and high dimensionality in data clustering must be considered.