Gaussian Mixture Models
PulseAugur coverage of Gaussian Mixture Models — every cluster mentioning Gaussian Mixture Models across labs, papers, and developer communities, ranked by signal.
4 day(s) with sentiment data
-
New algorithm drastically cuts GMM training time, enables billion-parameter models
Researchers have developed a novel variational approximation for Gaussian Mixture Models (GMMs) that significantly reduces computational complexity. This new algorithm scales linearly with data dimensionality and sublin…
-
New research details EM and MoM for Softmax Mixture Models
A new research paper explores the application of the Expectation-Maximization (EM) algorithm and the Method of Moments (MoM) to Softmax Mixture Models (SMMs). These models are used for analyzing probabilities in heterog…
-
New research details scaling laws for multiclass logistic regression training
A new paper published on arXiv details the training dynamics of multiclass logistic regression, establishing precise scaling laws for cross-entropy risk under gradient-based optimization. The research indicates that lea…
-
New SQS method achieves high DNN compression via Bayesian learning · 2 sources tracked
Researchers have developed a new method called SQS for compressing large neural networks, enabling their deployment on devices with limited resources. This unified framework simultaneously performs weight pruning and lo…
-
New research tackles Gaussian Mixture Model selection and estimation
Two new research papers explore advanced methods for selecting and estimating parameters in Gaussian Mixture Models (GMMs). The first paper focuses on one-dimensional GMMs, establishing optimal sampling complexity for m…
-
New Gauss-Hermite Quadrature Method for Gaussian Mixture Entropy
Researchers have developed a new Gauss-Hermite quadrature method to numerically approximate the differential entropy of Gaussian mixtures, which typically lacks a closed-form solution. This method's accuracy is influenc…
-
New amortized framework improves kernel density estimation bandwidth selection
Researchers have developed a novel amortized framework for learning bandwidth selection in kernel density estimation. This approach optimizes the logarithmic score across a distribution of density-estimation tasks, enab…
-
Gradient EM algorithm achieves global convergence for over-parameterized Gaussian Mixtures
Researchers have established a global convergence guarantee for the gradient Expectation-Maximization (EM) algorithm when applied to over-parameterized Gaussian Mixture Models (GMMs). This marks the first such result fo…
-
NLP pipeline detects accusatory language in Ecuador's public procurement
Researchers have developed a novel NLP pipeline to detect accusatory language in public procurement data from Ecuador's official system. This hybrid approach combines unsupervised clustering with supervised classificati…
-
New mathematical distances for partial optimal transport unveiled
Researchers have introduced new mathematical tools, entropic partial optimal transport and a partial mixture Gromov--Wasserstein distance, designed to compare probability measures and metric measure spaces. These method…
-
LLMs and GMMs Enhance NLP Clustering for Underrepresented Topics
Researchers have developed a new unsupervised data augmentation method for Natural Language Processing (NLP) that combines Gaussian Mixture Models (GMMs) and Large Language Models (LLMs). This approach aims to improve t…
-
New research explores Compactly Supported Radial Basis Functions for probability density modeling
Researchers have explored the use of Compactly Supported Radial Basis Functions (CS-RBFs) as a novel parametric family for probability density functions, particularly focusing on Wendland $\mathscr{C}^2$ kernels. The st…
-
Lloyd's K-Means Algorithm Identified as Special Case of Frank-Wolfe Method
A new paper establishes a connection between Lloyd's K-Means Clustering Algorithm and the Frank-Wolfe (FW) algorithm, demonstrating that K-Means is a specific instance of FW. This research derives a non-asymptotic conve…
-
New optimization framework uses Gaussian mixtures for robust chance-constrained problems
Researchers have developed a new method for distributionally robust linear chance-constrained problems, utilizing a Gaussian mixture model (GMM) to represent uncertainty. This approach improves upon finite-support distr…
-
Diffusion models' information processing and generative capabilities analyzed · 4 sources tracked
Recent research papers explore the inner workings of diffusion models, focusing on how they store and utilize information during the generative process. Studies indicate that these models commit significant information …
-
Self-Organizing Maps: An Underappreciated Clustering Algorithm
This article examines clustering algorithms, focusing on Self-Organizing Maps (SOMs) and their underappreciated potential. The author advocates for a deeper look into SOMs, suggesting that tuning them can yield signific…
-
New GMM-EVA framework enhances LVLM long video understanding
Researchers have introduced GMM-EVA, a novel framework designed to improve the efficiency and effectiveness of long video understanding in Large Vision-Language Models (LVLMs). This method utilizes Gaussian Mixture Mode…
-
New research explores Gaussian Mixture Models via spherical decomposition and statistical mechanics
Two new research papers explore Gaussian Mixture Models (GMMs) from different analytical perspectives. The first paper introduces a method using spherical radial decomposition to represent GMM probability functions as i…
-
New AI methods tackle time series forecasting and model explainability · 5 sources tracked
Researchers have introduced KARMA, a novel method for explaining time-series forecasting models by constructing a Markov surrogate model that captures temporal dependencies. This approach identifies the minimal history …
-
New method generates patient data for scarce medical AI training
Researchers have developed a novel patient augmentation technique for data-scarce medical Multiple Instance Learning (MIL). This method generates realistic patient data in embedding space by using Gaussian Mixture Model…