Designing and interoperability testing of distributed, application-level network protocols is complex. Windows, for example, supports currently more than 200 protocols, ranging fro...
Many existing privacy-preserving techniques for querying distributed databases of sensitive information do not scale for large databases due to the use of heavyweight cryptographi...
Sherman S. M. Chow, Jie-Han Lee, Lakshminarayanan ...
While phrase-based statistical machine translation systems currently deliver state-of-theart performance, they remain weak on word order changes. Current phrase reordering models ...
When constructing a classifier from labeled data, it is important not to assign too much weight to any single input feature, in order to increase the robustness of the classifier....
The residual variance estimation problem is well-known in statistics and machine learning with many applications for example in the field of nonlinear modelling. In this paper, we...