Heterogeneous information networks that contain multiple types of objects and links are ubiquitous in the real world, such as bibliographic networks, cyber-physical networks, and ...
Real-time surveillance systems, network and telecommunication systems, and other dynamic processes often generate tremendous (potentially infinite) volume of stream data. Effectiv...
Y. Dora Cai, David Clutter, Greg Pape, Jiawei Han,...
In this paper, we present an accurate and realtime PE-Miner framework that automatically extracts distinguishing features from portable executables (PE) to detect zero-day (i.e. pr...
M. Zubair Shafiq, S. Momina Tabish, Fauzan Mirza, ...
Ubiquitous Knowledge Discovery is a new research area at the intersection of machine learning and data mining with mobile and distributed systems. In this paper the main character...
Manifold learning is an effective methodology for extracting nonlinear structures from high-dimensional data with many applications in image analysis, computer vision, text data a...