Abstract--Many studies have shown the limits of support/confidence framework used in Apriori-like algorithms to mine association rules. One solution to cope with this limitation is...
Yannick Le Bras, Philippe Lenca, Sorin Moga, St&ea...
Hybrid learning methods use theoretical knowledge of a domain and a set of classified examples to develop a method for classification. Methods that use domain knowledge have been ...
Abstract—To achieve higher reliability, safety, and faulttolerance, many mission-critical detection and decision systems implement consensus algorithms that force the systems’ ...
k. The model we study can be interpreted as a broad, abstract extension of the well-studied on-line prediction model to a general decision-theoretic setting. We show that the multi...
— This paper presents a single-objective and a multiobjective stochastic optimization algorithms for global training of neural networks based on simulated annealing. The algorith...