Kernel-based tracking approaches have proven to be more efficient in computation compared to other tracking approaches such as particle filtering. However, existing kernel-based...
Record linkage is the problem of identifying similar records across different data sources. The similarity between two records is defined based on domain-specific similarity functi...
We present the Higher Order Proxy Neighborhoods (HOPS) approach to modeling higher order neighborhoods in Markov Random Fields (MRFs). HOPS incorporates more context information i...
The recursive least-squares (RLS) algorithm is one of the most well-known algorithms used in adaptive filtering, system identification and adaptive control. Its popularity is main...
High dimensionality remains a significant challenge for document clustering. Recent approaches used frequent itemsets and closed frequent itemsets to reduce dimensionality, and to...