Generalized Naive Bayes

In this paper we introduce the so-called Generalized Naive Bayes structure as an extension of the Naive Bayes structure. We give a new greedy algorithm that finds a good fitting Generalized Naive Bayes (GNB) probability distribution. We prove that this fits the data at least as well as the probability distribution determined by the classical Naive Bayes (NB). Then, under a not very restrictive condition, we give a second algorithm for which we can prove that it finds the optimal GNB probability distribution, i.e. best fitting structure in the sense of KL divergence. Both algorithms are constructed to maximize the information content and aim to minimize redundancy. Based on these algorithms, new methods for feature selection are introduced. We discuss the similarities and differences to other related algorithms in terms of structure, methodology, and complexity. Experimental results show, that the algorithms introduced outperform the related algorithms in many cases.

Matrix and graph representations of vine copula structures

Vine copulas can efficiently model a large portion of probability distributions. This paper focuses on a more thorough understanding of their structures. We are building on well-known existing constructions to represent vine copulas with graphs as well as matrices. The graph representations include the regular, cherry and chordal graph sequence structures, which we show equivalence between. Importantly we also show that when a perfect elimination ordering of a vine structure is given, then it can always be uniquely represented with a matrix. O. M. N\'apoles has shown a way to represent them in a matrix, and we algorithmify this previous approach, while also showing a new method for constructing such a matrix, through cherry tree sequences. Lastly, we prove that these two matrix-building algorithms are equivalent if the same perfect elimination ordering is being used.