K-means++
PulseAugur coverage of K-means++ — every cluster mentioning K-means++ across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
Sector-Mean Initialization offers faster, deterministic k-means clustering
Researchers have introduced Sector-Mean Initialization, a novel deterministic method for initializing centroids in k-means clustering. This approach partitions data into angular sectors around a global centroid and calc…
-
LLM crowds improved by behavioral clustering for future prediction
Researchers have developed a novel framework to improve the accuracy of future predictions made by large language models (LLMs). The approach focuses on creating diverse LLM crowds by clustering models based on their re…
-
K-means++ algorithm modified for improved approximation ratio
Researchers have proposed a modification to the k-means++ algorithm, a common method for initializing k-means clustering. The standard algorithm has a worst-case expected approximation ratio of \Theta(\log k) for a fixe…
-
New active learning strategy improves bioacoustic classification for rare calls
Researchers have developed a new active learning strategy called BADGE-Greedy-DPP for bioacoustic call-type classification, which is particularly effective for long-tailed and sparse datasets. This method greedily selec…
-
New PLSCAN algorithm offers improved multiscale density-based clustering
Researchers have introduced PLSCAN, a novel multiscale density-based clustering algorithm designed for exploratory data analysis. PLSCAN addresses the challenge of hyperparameter selection in existing density-based meth…
-
New criterion optimizes k-means++ restarts using data difficulty
Researchers have developed a new criterion called GTRC for the k-means++ algorithm to determine the optimal number of restarts. This method uses a Good-Turing estimate and confidence bounds to dynamically adjust restart…
-
AI framework identifies suspicious trading patterns using K-Means++ clustering
Researchers have developed a new toolkit using K-Means++ clustering to detect suspicious trading patterns in capital markets. The framework analyzes a dataset of approximately one million financial transactions from 201…
-
Lloyd's algorithm clustering consistency proven under perturbed samples
Researchers have analyzed the consistency of Lloyd's algorithm, a popular unsupervised clustering method, when applied to perturbed data. They demonstrated that even with small perturbations, the algorithm maintains an …