Title :
Privacy Preserving distributed structure learning of probabilistic graphical models
Author_Institution :
Dept. of Electr. Eng. & Comput. Sci., Univ. of Tennessee, Knoxville, TN, USA
Abstract :
Privacy preserving structure learning of probabilistic graphical model is studied using the framework of secure multi-party computation. Both constraint and score based learning procedures are rendered the capability of privacy preserving. A data set of adolescent health is used to learn the relationships related to drinking behaviors.
Keywords :
data privacy; distributed processing; health care; learning (artificial intelligence); probability; adolescent health; constraint procedure; drinking behaviors; privacy preserving distributed structure learning; probabilistic graphical models; score based learning procedure; secure multiparty computation; Bayes methods; Conferences; Data privacy; Graphical models; Privacy; Random variables;
Conference_Titel :
Globecom Workshops (GC Wkshps), 2013 IEEE
Conference_Location :
Atlanta, GA
DOI :
10.1109/GLOCOMW.2013.6824984
