A Family of Kernelized Matrix Costs for Multiple-Output Mixture Neural Networks

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Hu, Bo, Príncipe, José C.
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918156183797760
author Hu, Bo
Príncipe, José C.
author_facet Hu, Bo
Príncipe, José C.
contents Pairwise distance-based costs are crucial for self-supervised and contrastive feature learning. Mixture Density Networks (MDNs) are a widely used approach for generative models and density approximation, using neural networks to produce multiple centers that define a Gaussian mixture. By combining MDNs with contrastive costs, this paper proposes data density approximation using four types of kernelized matrix costs in the Hilbert space: the scalar cost, the vector-matrix cost, the matrix-matrix cost (the trace of Schur complement), and the SVD cost (the nuclear norm), for learning multiple centers required to define a mixture density.
format Preprint
id arxiv_https___arxiv_org_abs_2509_24076
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Family of Kernelized Matrix Costs for Multiple-Output Mixture Neural Networks
Hu, Bo
Príncipe, José C.
Machine Learning
Pairwise distance-based costs are crucial for self-supervised and contrastive feature learning. Mixture Density Networks (MDNs) are a widely used approach for generative models and density approximation, using neural networks to produce multiple centers that define a Gaussian mixture. By combining MDNs with contrastive costs, this paper proposes data density approximation using four types of kernelized matrix costs in the Hilbert space: the scalar cost, the vector-matrix cost, the matrix-matrix cost (the trace of Schur complement), and the SVD cost (the nuclear norm), for learning multiple centers required to define a mixture density.
title A Family of Kernelized Matrix Costs for Multiple-Output Mixture Neural Networks
topic Machine Learning
url https://arxiv.org/abs/2509.24076