We prove that if NP ⊆ BPP, i.e., if SAT is worst-case hard, then for every probabilistic polynomial-time algorithm trying to decide SAT, there exists some polynomially samplable ...
Abstract. One of the maincharacteristics of logical reasoning in knowledge based systems is its high computational complexity. Anytime deduction and anytime compilation are two att...
Abstract. Kernel based methods (such as k-nearest neighbors classifiers) for AI tasks translate the classification problem into a proximity search problem, in a space that is usu...
Bayesian networks are directed acyclic graphs that represent dependencies between variables in a probabilistic model. Many time series models, including the hidden Markov models (H...
This article proposes a novel similarity measure between
vector sequences. Recently, a model-based approach was
introduced to address this issue. It consists in modeling
each se...