In this paper, we present a unified framework for reduced space modeling and rendering of dynamic and nonhomogenous participating media, like snow, smoke, dust and fog. The key id...
Abstract. We describe an algorithm for fitting a Catmull-Clark subdivision surface model to an unstructured, incomplete and noisy data set. We complete the large missing data regi...
Decision-tree algorithms are known to be unstable: small variations in the training set can result in different trees and different predictions for the same validation examples. B...
Intrusion detection systems (IDSs) fall into two high-level categories: network-based systems (NIDS) that monitor network behaviors, and host-based systems (HIDS) that monitor sys...
Clustering methods can be either data-driven or need-driven. Data-driven methods intend to discover the true structure of the underlying data while need-driven methods aims at org...