Mixtures, envelopes and hierarchical duality
نویسندگان
چکیده
منابع مشابه
Mixtures, envelopes, and hierarchical duality
We develop a connection between mixture and envelope representations of objective functions that arise frequently in statistics. We refer to this connection using the term “hierarchical duality.” Our results suggest an interesting and previously underexploited relationship between marginalization and profiling, or equivalently between the Fenchel–Moreau theorem for convex functions and the Bern...
متن کاملDistributive Envelopes and Topological Duality for Lattices via Canonical Extensions
We establish a topological duality for bounded lattices. The two main features of our duality are that it generalizes Stone duality for bounded distributive lattices, and that the morphisms on either side are not the standard ones. A positive consequence of the choice of morphisms is that those on the topological side are functional. Towards obtaining the topological duality, we develop a unive...
متن کاملInstrument Recognition in Polyphonic Mixtures Using Spectral Envelopes
Instrument recognition in polyphonic music is a difficult task in computer audition. Many current methods approach this problem by first attempting to separate the timbre features among the sources present in the mixture using source separation, multi-pitch estimation, or noteonset techniques. Instrument (timbre) recognition then proceeds on these separated features. This study proposes another...
متن کاملBayesian Hierarchical Mixtures of Experts
The Hierarchical Mixture of Experts (HME) is a well-known tree-structured model for regression and classification, based on soft probabilistic splits of the input space. In its original formulation its parameters are determined by maximum likelihood, which is prone to severe overfitting, including singularities in the likelihood function. Furthermore the maximum likelihood framework offers no n...
متن کاملGrowing Hierarchical Mixtures of Experts
We propose a novel approach to automatically growing and pruning Hierarchical Mixtures of Experts. The constructive algorithm proposed here enables large hierarchies consisting of several hundred experts to be trained e ectively. We show that HME's trained by our automatic growing procedure yield better generalization performance than traditional static and balanced hierarchies. Evaluation of t...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Journal of the Royal Statistical Society: Series B (Statistical Methodology)
سال: 2015
ISSN: 1369-7412,1467-9868
DOI: 10.1111/rssb.12130