iCVI-ARTMAP: Using Incremental Cluster Validity Indices and Adaptive Resonance Theory Reset Mechanism to Accelerate Validation and Achieve Multiprototype Unsupervised Representations

Silva, Leonardo Enzo Brito da; Rayapati, Nagasharath; Wunsch, Donald C.

doi:10.1109/tnnls.2022.3160381

Cited by 11 publications

(7 citation statements)

References 82 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Indeed, methods have been proposed which focus solely on the proper selection of a CVI for optimal results on a particular data set [19]. However, an extension of [4] found that a single iCVI can serve as a vigilance mechanism for Fuzzy ART [20] as well as for TopoARTMAP [21]. The former study investigated iCH, iWB, iXB, iDB, incremental Pakhira-Bandyopadhyay-Maulik (iPBM), and incremental Negentropy Increment (iNI).…”

Section: Background and Related Work A Clusteringmentioning

confidence: 99%

Meta-iCVI: Ensemble Validity Metrics for Concise Labeling of Correct, Under- or Over-Partitioning in Streaming Clustering

Melton,

Petrenko,

Wunsch

2024

IEEE Access

Self Cite

View full text Add to dashboard Cite

Understanding the performance and validity of clustering algorithms is both challenging and crucial, particularly when clustering must be done online. Until recently, most validation methods have relied on batch calculation and have required considerable human expertise in their interpretation. Improving real-time performance and interpretability of cluster validation, therefore, continues to be an important theme in improving unsupervised learning. Building upon previous work on incremental cluster validity indices (iCVIs), this paper introduces the Meta-iCVI as a tool for explainable and concise labeling of partition quality in online clustering. Past work has shown that some iCVIs are better at detecting under-partition; others at over-partition. Combining them was hypothesized to improve cluster validation analysis. Experiments were conducted on generalized synthetic and real-world data sets to demonstrate the efficacy and application of this method. Results of 100% accuracy were achieved in labeling partition quality on real-world data sets including MNIST and FLIR ADAS, demonstrating that the Meta-iCVI is a powerful and efficient tool for classifying partition quality in a variety of conditions. Its introduction should empower new and more efficient streaming clustering techniques. Additionally, we believe this to be the first implementation of an ensemble iCVI metric and the first time iCVI validation performance has been evaluated on randomized sample presentation.

show abstract

Section: Background and Related Work A Clusteringmentioning

confidence: 99%

Meta-iCVI: Ensemble Validity Metrics for Concise Labeling of Correct, Under- or Over-Partitioning in Streaming Clustering

Melton,

Petrenko,

Wunsch

2024

IEEE Access

Self Cite

View full text Add to dashboard Cite

show abstract

“…Though these algorithms in large part belong to the class of incremental neurogenesis clustering/unsupervised algorithms, they have been adapted for applications in supervised, reinforcement, and even multimodal learning [19,39], tackling clustering issues from sample granularity [25,26] to distributed representations [40][41][42], pattern sequences [43], context recognition [44,45], and uncertainties [46][47][48]. Some algorithms based upon ART have even been combined with incremental cluster validity indices (ICVIs), metrics of clustering performance in the absence of supervised labels, to enable a variety of incremental, online, and multimodal clustering and biclustering applications [49][50][51][52][53][54]. ART algorithms are additionally well suited for lifelong learning (L2) applications because they are derived from theories on how biological neural networks address the stability-plasticity dilemma to mitigate catastrophic forgetting [55][56][57].…”

Section: Adaptive Resonance Theorymentioning

confidence: 99%

Analyzing Biomedical Datasets with Symbolic Tree Adaptive Resonance Theory

Petrenko,

Hier,

Bone

et al. 2024

Information

View full text Add to dashboard Cite

Biomedical datasets distill many mechanisms of human diseases, linking diseases to genes and phenotypes (signs and symptoms of disease), genetic mutations to altered protein structures, and altered proteins to changes in molecular functions and biological processes. It is desirable to gain new insights from these data, especially with regard to the uncovering of hierarchical structures relating disease variants. However, analysis to this end has proven difficult due to the complexity of the connections between multi-categorical symbolic data. This article proposes symbolic tree adaptive resonance theory (START), with additional supervised, dual-vigilance (DV-START), and distributed dual-vigilance (DDV-START) formulations, for the clustering of multi-categorical symbolic data from biomedical datasets by demonstrating its utility in clustering variants of Charcot–Marie–Tooth disease using genomic, phenotypic, and proteomic data.

show abstract

“…Indeed, methods have been proposed which focus solely on the proper selection of an CVI for optimal results on a particular data set [13]. However, an extension of [4] found that a single iCVI can serve as a vigilance mechanism for Fuzzy ART [14] as well as for TopoARTMAP [15]. The former study investigated iCH, iWB, iXB, iDB, incremental Pakhira-Bandyopadhyay-Maulik (iPBM), and incremental Negentropy Increment (iNI).…”

Section: Background and Related Work A Clusteringmentioning

confidence: 99%

Meta-iCVIs: Ensemble Validity Metrics for Concise Labeling of Correct, Under- or Over-Partitioning in Streaming Clustering

Petrenko¹,

Melton²,

Wunsch³

2023

Preprint

View full text Add to dashboard Cite

<p>Understanding the performance and validity of clustering algorithms is both challenging and crucial, particularly when clustering must be done online. Until recently, most validation methods have relied on batch calculation and have required considerable human expertise in their interpretation. Improving real-time performance and interpretability of cluster validation, therefore, continues to be an important theme in improving unsupervised learning. Building upon previous work on incremental cluster validity indices (iCVIs), this paper introduces the Meta-iCVI as a tool for explainable and concise labeling of partition quality in online clustering. Some iCVIs are better at detecting under-partition; others at over-partition. Combining them was hypothesized to improve cluster validation analysis. Experiments were conducted on generalized synthetic and real-world data sets to demonstrate the efficacy and application of this method.</p> <p>Results of 100% accuracy were achieved in labeling partition quality on real-world data sets including MNIST and FLIR ADAS, demonstrating that the Meta-iCVI is a powerful and efficient tool for classifying partition quality in a variety of conditions. Its introduction should empower new and more efficient streaming clustering techniques. Additionally, we believe this to be the first implementation of an ensemble iCVI metric and the first time iCVI validation performance has been evaluated on randomized sample presentation.</p>

show abstract

iCVI-ARTMAP: Using Incremental Cluster Validity Indices and Adaptive Resonance Theory Reset Mechanism to Accelerate Validation and Achieve Multiprototype Unsupervised Representations

Cited by 11 publications

References 82 publications

Meta-iCVI: Ensemble Validity Metrics for Concise Labeling of Correct, Under- or Over-Partitioning in Streaming Clustering

Meta-iCVI: Ensemble Validity Metrics for Concise Labeling of Correct, Under- or Over-Partitioning in Streaming Clustering

Analyzing Biomedical Datasets with Symbolic Tree Adaptive Resonance Theory

Meta-iCVIs: Ensemble Validity Metrics for Concise Labeling of Correct, Under- or Over-Partitioning in Streaming Clustering

Contact Info

Product

Resources

About