A novel self-directed learning framework for cluster ensemble

Kadhim, Mustafa R.; Tian, Wenhong; Zhou, Guangyao

doi:10.1016/j.jksuci.2022.07.003

Cited by 4 publications

(8 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recent studies have explored the GSA and discovered several drawbacks like memory-less, weak convergence velocity, local optima trapping, and fuzzy solution precision. Further, Equations ( 2) and ( 6) use randomization, which has been proved invalid unless there is a powerful detection model [6]. Many recent models supported several solutions by advancing the positioning, velocity, and Kbest equations.…”

Section: Related Work a Gravitational Search Algorithmmentioning

confidence: 99%

“…Constraints issues such as weak or violent constraints result in bad performance or termination execution. Some models reached overstated resources because of constraints or technical issues [6], [42], [43].…”

Section: The Otc Issuesmentioning

confidence: 99%

“…The SSLbased clustering clusters the given data based on the provided side information (constraints) to improve the prediction process of the model. Also, recent developments were not limited to SSL, where many researchers introduced novel frameworks like ensemble [5], self-directed [6], non-parametric [7], and graph representation [8]. These integrations were introduced to sort out the traditional drawbacks and improve performance.…”

Section: Introductionmentioning

confidence: 99%

“…These integrations were introduced to sort out the traditional drawbacks and improve performance. However, recent studies investigated and proved that "Semisupervised clustering (SSC) may be highly inconsistent with the true partition" [1], which is very obvious in the investigation done by [6], where the author showed many cluster-based models are incompetent in long-term real-world applications (LTRWA).…”

Section: Introductionmentioning

confidence: 99%

“…For instance, a single model performs in a single domain but fails to work with others. Second, the growing data concern has been proven by [6] that many clustering models are memory-less and need to re-determine the centroids and cluster the whole data in case new data is inserted. Thus, their experimental results for the baseline models are suffering from the same issue.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Gravity Sub-centroids for Optimal Clustering

Mustafa,

Qingyuan,

Jianbo

et al. 2023

Preprint

View full text Add to dashboard Cite

This work highlights issues that were not deeply investigated in previous studies on clustering solutions, which have essential impacts on performance in long-term real-world applications that are challenging to detect instantly. Thus, we addressed these issues by proposing two novel techniques: first, we expand the idea of clustering based on centroids to multiple sub-centroids that assist assignment functions in finding the optimal solution. In contrast to recent studies, we extended the concept of gravitational force toward clustering solutions. Finally, the introduced gap generation concept has been associated with these techniques to support a superior clustering solution. Our model is termed semi-supervised gravity clustering (SSGC). To demonstrate the strength of SSGC, we consider multiple performance measurements besides the traditional ones to validate the clustering models in various scenarios. The experimental results show that SSGC outperforms baseline models and successfully obtains the best performance of 30 different domain datasets. Finally, our methodology code is already released.

show abstract

Section: Related Work a Gravitational Search Algorithmmentioning

confidence: 99%