Density-Based Clustering Based on Hierarchical Density Estimates

By Ricardo José Gabrielli Barreto Campello, Davoud Moulavi, Jörg Sander. Pacific-Asia Conference on Knowledge Discovery and Data Mining, 2013

2,572 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Signal Processing

View Paper: https://www.semanticscholar.org/paper/706bfdaffb2c66415ca9fe74edf574d2d0476f59

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Signal Processing

Citations in the corpus

Study on density peaks clustering based on k-nearest neighbors and principal component analysis (2016, methodology): “Density-based clustering [3–7] is represented by DBSCAN [3] .”
Obesity shapes metabolism in the tumor microenvironment to suppress anti-tumor immunity (2020, methodology): “To identify major cell populations, dimensional reduction using t-Distributed Stochastic Neighbor Embedding (t-SNE) was performed on the cleaned dataset followed by density-based clustering by HDBSCAN (Figure 4B) (Campello et al., 2013; van der Maaten and Hinton, 2008).”
Finding, visualizing, and quantifying latent structure across diverse animal vocal repertoires (2020, methodology): “We compared four different labeling techniques: a hierarchical density-based clustering algorithm (HDBSCAN; (Campello et al., 2013; McInnes et al., 2017)) applied to UMAP projections of spectrograms, HDBSCAN applied to PCA projections of spectrograms1, k-means (Pedregosa et al.”
DBSCAN Revisited, Revisited (2017, methodology): “Publication date: July 2017. such as OPTICS [3], and HDBSCAN* [13], which remove the need to choose ε .”
Machine learning for Internet of Things data analysis: A survey (2017, background): “Additionally, the performance of the model is highly sensitive to the distance metric that is used for determining if a region is dense [81].”
Density-Based Clustering Validation (2014, result): “RJGBC thanks CNPq (grant #301241 2010-4) and FAPESP (grants#2010/20032−6 and#2013/18698-4).”
Do travelers' reviews depend on the destination? An analysis in coastal and urban peer‐to‐peer lodgings (2021, methodology): “20d‐UMAP embedding is clustered with Hierarchical Density‐Based Spatial Clustering of Applications with Noise algorithm (from now on, hDBSCAN; Campello et al., 2013; McInnes & Healy, 2017).”
Interpreting Word-Level Hidden State Behaviour of Character-Level LSTM Language Models (2018, methodology): “We chose to use the HDBSCAN clustering algorithm (Campello et al., 2013), since it is designed to work with non-globular clusters of varying density, does not require that an expected number of clusters be specified in advance, and is willing to avoid assigning points to a cluster if they do not”
Top2Vec: Distributed Representations of Topics (2020, methodology): “In order ﬁnd the dense areas of documents in the semantic space, density based clustering is used on the document vectors, speciﬁcally Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN) [25, 26, 27].”
Density‐based clustering (2019, background): “Campello, Moulavi, and Sander (2013) and Campello et al. (2015) show that by performing a simple transformation of the data from the original space to a transformed space of pairwise dissimilarities, called mutual reachability distances, all DBSCAN clustering solutions for all possible values of ε…”
Distribution Density, Tails, and Outliers in Machine Learning: Metrics and Applications (2019, methodology): “We apply t-SNE (Maaten & Hinton, 2008) on the pixel space (for MNIST and Fashion-MNIST) or ResNetv2 feature space (for CIFAR10) to project the example sets into two dimensions, and cluster with HDBSCAN (Campello et al., 2013), a hierarchical and density-based clustering algorithm which does not try ”
Exploring Backdoor Poisoning Attacks Against Malware Classifiers (2020, neutral): “Additionally, HDBSCAN does not require a fixed number of clusters, but has two other parameters that control the cluster density (minimum size of a cluster, set at 1% of the training benign data, 3000 points, and minimum number of samples to form a dense region, set at 0.5%, 600 points).”
Hierarchical Density Estimates for Data Clustering, Visualization, and Outlier Detection (2015, methodology): “In the following, we define the density-based clustering algorithm DBSCAN* as in our preliminary work [Campello et al. 2013a], which differs from DBSCAN [Ester et al. 1996] in that the clusters are defined based on core objects alone.”
Dark Patterns at Scale (2019, methodology): “ture at least 95% of the variance in the data, resulting in 3 and 10 components, respectively. We used the Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN) algorithm [34] implemented in the HDBSCAN Python library [14] to extract clusters from this data. We chose HDB”
Structural and functional characterization of G protein–coupled receptors with deep mutational scanning (2019, methodology): “Each residue is assigned into one of six clusters using HDBSCAN (see Fig.”
FLAME: Taming Backdoors in Federated Learning (2022, methodology): “Then HDBSCAN uses the stability value to determine the cluster candidate with the most homogeneous density and returns this candidate as majority cluster.”
3D particle shape modelling and optimization through proper orthogonal decomposition (2017, positive): “An optimal choice, minimizing the sensitivity of the obtained set of clusters to variations of in some range about the set value, is identified on implementing the HDBSCAN algorithm [48].”