Journal article · 2026 · Open access
Central description length (CDL) clustering validation index
Artificial Intelligence, 361, 104644 · Published 20 October 2026 · DOI 10.1016/j.artint.2026.104644
In brief
“Central description length (CDL) clustering validation index” is a 2026 journal article by Mahdi Shamsi and Soosan Beheshti, published in Artificial Intelligence. OpenAlex counts 0 citations as of 17 November 2026. It is open access under a CC BY-NC licence: the full text is free to read at Artificial Intelligence.
- Citations
- 0
- References
- 52
- Authors
- 2
Abstract
Clustering is an unsupervised machine-learning technique that groups unlabeled data and is used in various applications to provide valuable insights into underlying patterns and structures within datasets. A critical issue in unsupervised clustering is choosing the best algorithm among a set of clustering algorithms with the optimal set of hyperparameters for a given dataset. Clustering validation index (CVI) approaches aim to provide the optimal tuning hyperparameter and compare different clustering algorithms. Existing CVI methods rely on the convexity of the datasets and do not perform well for non-convex and irregular datasets. In these scenarios, additional kernels are used in the clustering process that introduces computational complexity. This paper presents a new CVI approach denoted by Central Description Length (CDL). The method utilizes the cluster compactness to provide the description length of the true clusters based on the clustering technique. The approach enables evaluation and comparison of the clustering algorithm hyperparameters without any assumption on the data convexity. Consequently, the method is robust to the data structure and if possible, can eliminate the extra burden of using the kernels by optimally tuning the hyperparameters of the clustering method itself. Distinct from conventional approaches, the proposed method utilizes the inherent data structure to precisely determine separation and scatteredness, manifesting effective handling of both non-convex datasets and convex clusters that are closely positioned or overlapping. Simulation results for challenging datasets with arbitrary shapes demonstrate the superiority of the proposed CDL-CVI compared to the existing CVIs in detecting the true number of clusters and in the sense of Adjusted Rand Index (ARI) and Normalized Variation of Information (NVI). In addition, the CDL-CVI was evaluated on large scale real world dataset with high dimensional features. The results demonstrate the robustness and scalability of CDL-CVI which confirms superiority of the proposed approach compared to existing clustering validation indices in terms of ARI and NVI.
Cite this paper
Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.
APA 7
Shamsi, M., & Beheshti, S. (2026). Central description length (CDL) clustering validation index. Artificial Intelligence, 361, 104644. https://doi.org/10.1016/j.artint.2026.104644
MLA 9
Shamsi, Mahdi, and Soosan Beheshti. “Central description length (CDL) clustering validation index.” Artificial Intelligence, vol. 361, 2026, p. 104644. https://doi.org/10.1016/j.artint.2026.104644.
Chicago
Shamsi, Mahdi and Soosan Beheshti. 2026. “Central description length (CDL) clustering validation index.” Artificial Intelligence 361: 104644. https://doi.org/10.1016/j.artint.2026.104644.
Harvard
Shamsi, M. and Beheshti, S. (2026) ‘Central description length (CDL) clustering validation index’, Artificial Intelligence, 361, p. 104644. doi:10.1016/j.artint.2026.104644.
IEEE
M. Shamsi and S. Beheshti, “Central description length (CDL) clustering validation index,” Artificial Intelligence, vol. 361, p. 104644, 2026, doi: 10.1016/j.artint.2026.104644.
BibTeX
@article{shamsi2026central,
title = {{Central description length (CDL) clustering validation index}},
author = {Shamsi, Mahdi and Beheshti, Soosan},
journal = {Artificial Intelligence},
year = {2026},
volume = {361},
pages = {104644},
doi = {10.1016/j.artint.2026.104644},
url = {https://doi.org/10.1016/j.artint.2026.104644}
}
Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)
Topic and keywords
Physical Sciences › Computer Science › Artificial Intelligence › Advanced Clustering Algorithms Research
- cluster compactness
- hyperparameter optimization
- clustering algorithm comparison
- high-dimensional data
Work with this paper
Most-cited papers in Computer Science
- ImageNet classification with deep convolutional neural networksAlex Krizhevsky, Ilya Sutskever, Geoffrey E. Hinton · Communications of the ACM · 2017 · 114,726 citations
- Long Short-Term MemorySepp Hochreiter, Jürgen Schmidhuber · Neural Computation · 1997 · 109,097 citations
- A Mathematical Theory of CommunicationClaude E. Shannon · Bell System Technical Journal · 1948 · 91,736 citations
- Deep learningYann LeCun, Yoshua Bengio, Geoffrey E. Hinton · Nature · 2015 · 90,226 citations
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal NetworksShaoqing Ren, Kaiming He, Ross Girshick and 1 more · IEEE Transactions on Pattern Analysis and Machine Intelligence · 2016 · 74,143 citations
Questions about this paper
How do I cite this paper in APA style?
Shamsi, M., & Beheshti, S. (2026). Central description length (CDL) clustering validation index. Artificial Intelligence, 361, 104644. https://doi.org/10.1016/j.artint.2026.104644
How many times has this paper been cited?
0 times, according to OpenAlex on 17 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.
Is this paper free to read?
Yes. It is open access at Artificial Intelligence: sign in and open the full text from this page.
Who wrote this paper?
Mahdi Shamsi, Soosan Beheshti.
What is the DOI of this paper?
10.1016/j.artint.2026.104644, permanently at https://doi.org/10.1016/j.artint.2026.104644. Sign in and open it from this page.
Bibliographic data from OpenAlex (CC0), last checked 17 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.