Learning Bayesian network structure from large-scale data sets, without any expert-specified ordering of variables, remains a difficult problem. We propose systematic improvements to automatically learn Bayesian network structure from data. (1) We propose a linear parent search method to generate candidate graph. (2) We propose a comprehensive approach to eliminate cycles using minimal likeliho...

Web sites and transactional databases collect large amounts of time-stamped data. Businesses often want to make future predictions (forecasts) based on numerous sets of time-stamped data (sets of transactions). There are many time series analysis techniques related to forecasting, and an experienced analyst can effectively use these techniques to analyze, model, and forecast time series data. H...

Monte-Carlo path tracing algorithms for computer graphics require that given an incident light ray at a surface an outgoing direction can be computed with a distribution given by the magnitude of the bidirectional reeectance distribution function (BRDF). For analytic reeectance functions this can be done using various techniques including inverting the function, or tabulating some representatio...

Path Ranking Algorithm (PRA) addresses classification and retrieval tasks using learned combinations of labeled paths through a graph. Unlike most Statistical Relational Learning (SRL) methods, PRA scales to large data sets but uses a limited set of paths in its models—ones that correspond to short first order rules with no constants. We consider extending PRA in two ways—learning paths that co...

In this paper, we propose a path relinking procedure for the fixed-charge capacitated multicommodity network design problem. Cycle-based neighbourhoods are used both to move along path between elite solutions and to generate the elite candidate set by a tabu-like local search procedure. Several variants of the method are implemented and compared. Extensive computational experiments indicate tha...

Frequent pattern mining is a process of extracting frequently occurring itemset patterns from very large data storages. These frequent patterns are used to generate association rules which define the relationship among items. The strength of the relationship can be measured using two different units namely support value and confidence level. Any relationship that satisfies minimum threshold of ...

The native representation of LDA-style topics is a multinomial distributions over words, but automatic labelling of such topics has been shown to help readers interpret the topics better. We propose a novel framework for topic labelling using word vectors and letter trigram vectors. We generate labels automatically and propose automatic and human evaluations of our method. First, we use a chunk...

A shortest path problem is a practical issue in networks for real-world situations. This paper addresses the fuzzy shortest path (FSP) problem to obtain the best fuzzy path among fuzzy paths sets. For this purpose, a new efficient algorithm is introduced based on a new definition of ideal fuzzy sets (IFSs) in order to determine the fuzzy shortest path. Moreover, this algorithm is developed for ...