Toward Multidiversified Ensemble Clustering of High-Dimensional Data: From Subspaces to Metrics and Beyond

Multi-view clustering by CPS-merge analysis with application to multimodal single-cell data

Multi-view data can be generated from diverse sources, by different technologies, and in multiple modalities. In various fields, integrating information from multi-view data has pushed the frontier of discovery. In this paper, we develop a new approach for multi-view clustering, which overcomes the limitations of existing methods such as the need of pooling data across views, restrictions on the clustering algorithms allowed within each view, and the disregard for complementary information between views. Our new method, called CPS-merge analysis, merges clusters formed by the Cartesian product of single-view cluster labels, guided by the principle of maximizing clustering stability as evaluated by CPS analysis. In addition, we introduce measures to quantify the contribution of each view to the formation of any cluster. CPS-merge analysis can be easily incorporated into an existing clustering pipeline because it only requires single-view cluster labels instead of the original data. We can thus readily apply advanced single-view clustering algorithms. Importantly, our approach accounts for both consensus and complementary effects between different views, whereas existing ensemble methods focus on finding a consensus for multiple clustering results, implying that results from different views are variations of one clustering structure. Through experiments on single-cell datasets, we demonstrate that our approach frequently outperforms other state-of-the-art methods.

An Ensemble and Multi-View Clustering Method Based on Kolmogorov Complexity

The ability to build more robust clustering from many clustering models with different solutions is relevant in scenarios with privacy-preserving constraints, where data features have a different nature or where these features are not available in a single computation unit. Additionally, with the booming number of multi-view data, but also of clustering algorithms capable of producing a wide variety of representations for the same objects, merging clustering partitions to achieve a single clustering result has become a complex problem with numerous applications. To tackle this problem, we propose a clustering fusion algorithm that takes existing clustering partitions acquired from multiple vector space models, sources, or views, and merges them into a single partition. Our merging method relies on an information theory model based on Kolmogorov complexity that was originally proposed for unsupervised multi-view learning. Our proposed algorithm features a stable merging process and shows competitive results over several real and artificial datasets in comparison with other state-of-the-art methods that have similar goals.

Secuer: Ultrafast, scalable and accurate clustering of single-cell RNA-seq data

Identifying cell clusters is a critical step for single-cell transcriptomics study. Despite the numerous clustering tools developed recently, the rapid growth of scRNA-seq volumes prompts for a more (computationally) efficient clustering method. Here, we introduce Secuer, a Scalable and Efficient speCtral clUstERing algorithm for scRNA-seq data. By employing an anchor-based bipartite graph representation algorithm, Secuer enjoys reduced runtime and memory usage over one order of magnitude for datasets with more than 1 million cells. Meanwhile, Secuer also achieves better or comparable accuracy than competing methods in small and moderate benchmark datasets. Furthermore, we showcase that Secuer can also serve as a building block for a new consensus clustering method, Secuer-consensus, which again improves the runtime and scalability of state-of-the-art consensus clustering methods while also maintaining the accuracy. Overall, Secuer is a versatile, accurate, and scalable clustering framework suitable for small to ultra-large single-cell clustering tasks.

A Normal Distributed Dwarf Mongoose Optimization Algorithm for Global Optimization and Data Clustering Applications

As data volumes have increased and difficulty in tackling vast and complicated problems has emerged, the need for innovative and intelligent solutions to handle these difficulties has become essential. Data clustering is a data mining approach that clusters a huge amount of data into a number of clusters; in other words, it finds symmetric and asymmetric objects. In this study, we developed a novel strategy that uses intelligent optimization algorithms to tackle a group of issues requiring sophisticated methods to solve. Three primary components are employed in the suggested technique, named GNDDMOA: Dwarf Mongoose Optimization Algorithm (DMOA), Generalized Normal Distribution (GNF), and Opposition-based Learning Strategy (OBL). These parts are used to organize the executions of the proposed method during the optimization process based on a unique transition mechanism to address the critical limitations of the original methods. Twenty-three test functions and eight data clustering tasks were utilized to evaluate the performance of the suggested method. The suggested method’s findings were compared to other well-known approaches. In all of the benchmark functions examined, the suggested GNDDMOA approach produced the best results. It performed very well in data clustering applications showing promising performance.

Ensemble clustering of longitudinal bivariate HIV biomarker profiles to group patients by patterns of disease progression

This paper describes an ensemble cluster analysis of bivariate profiles of HIV biomarkers, viral load and CD4 cell counts, which jointly measure disease progression. Data are from a prevalent cohort of HIV positive participants in a clinical trial of vitamin supplementation in Botswana. These individuals were HIV positive upon enrollment, but with unknown times of infection. To categorize groups of participants based on their patterns of progression of HIV infection using both biomarkers, we combine univariate shape-based cluster results for multiple biomarkers through the use of ensemble clustering methods. We first describe univariate clustering for each of the individual biomarker profiles, and make use of shape-respecting distances for clustering the longitudinal profile data. In our data, profiles are subject to either missing or irregular measurements as well as unobserved initiation times of the process of interest. Shape-respecting distances that can handle such data issues, preserve time-ordering, and identify similar profile shapes are useful in identifying patterns of disease progression from longitudinal biomarker data. However, their performance with regard to clustering differs by severity of the data issues mentioned above. We provide an empirical investigation of shape-respecting distances (Fréchet and dynamic time warping (DTW)) on benchmark shape data, and use DTW in cluster analysis of biomarker profile observations. These reveal a primary group of ‘typical progressors,’ as well as a smaller group that shows relatively rapid progression. We then refine the analysis using ensemble clustering for both markers to obtain a single classification. The information from joint evaluation of the two biomarkers combined with ensemble clustering reveals subgroups of patients not identifiable through univariate analyses; noteworthy subgroups are those that appear to represent recently and chronically infected subsets.