High-dimensional Research Articles

Fast generation of high-dimensional spatial extremes

Widespread extreme climate events cause many fatalities, economic losses and have a huge impact on critical infrastructure. It is therefore of utmost importance to estimate the frequency and associated consequences of spatially concurrent extremes. Impact studies of climate extremes are severely hampered by the lack of extreme observations, and even large ensembles of climate simulations often do not include enough extreme or record-breaking climate events for robust analysis. On the other hand, weather generators specifically fitted to extreme observations can quickly generate many physically or statistically plausible extreme events, even with intensities that have never been observed before. We propose a Fourier-based algorithm for generating high-resolution synthetic datasets of rare events, using essential concepts of classical modelling of (spatial) extremes. Here, the key feature is that the stochastically generated datasets have the same spatial dependence as the observed extreme events. Using high-resolution gridded precipitation and temperature datasets, we show that the new algorithm produces realistic spatial patterns, and is particularly attractive compared to other existing methods for spatial extremes. It is exceptionally fast, easy to implement, scalable to high dimensions and, in principle, applicable for any spatial resolution. We generated datasets with 10000 gridpoints, a number that can be increased without difficulty. Since current impact models often require high-resolution climate inputs, the new algorithm is particularly useful for improved impact and vulnerability assessment.

Fault diagnoses of a nonlinear cracked rotor-bearing system based on vibration energy space and incremental learning approach

Health services of rotor-bearing systems are directly related to safe and stable works of rotating machineries in the industrial production. To address the issues of a low impact on vibration characteristics caused by faults in the higher rotational speed region, narrow and weak excitation signals, high dimensionality related to fault induction mechanisms, and difficulty in diagnosing using artificial intelligence technology due to the small number of fault samples in rotor systems, a novel fault detection method of nonlinear rotor systems through the incremental two-dimensional principal component analysis (I2DPCA) of the machine learning on the vibration energy space is proposed. Unlike mainly analyzing dynamic characteristics of rotor-bearing systems in vibration space, the method applies signals of the vibration energy space such as energy tracks, energy-time histories and energy-frequency spectra to have more advantages in vibration amplitudes caused by cracks in the higher rotational speed region and to overcome narrow and weak signals in the early fault stage. Then, the I2DPCA is applied to extract online low dimensional fault features and to process small vibration features of the rotor system. Simulation and experiment results show that performances of between-class margins and within-class clusters of different health conditions on the vibration energy space are more perfect than that on the vibration space. Based on the I2DPCA algorithm and the energy space method, the recognition rate of crack faults within the high rotational speed region is up to 99.6%, and it has good convergence and online learning ability in the case of small samples. The achieved conclusions of the method could provide a novel reference direction for fault diagnoses of rotor systems.

Interpretable machine learning uncovers epithelial transcriptional rewiring and a role for Gelsolin in COPD.

Transcriptomic analyses have advanced the understanding of complex disease pathophysiology including chronic obstructive pulmonary disease (COPD). However, identifying relevant biologic causative factors has been limited by the integration of high dimensionality data. COPD is characterized by lung destruction and inflammation with smoke exposure being a major risk factor. To define novel biological mechanisms in COPD, we utilized unsupervised and supervised interpretable machine learning analyses of single cell-RNA sequencing data from the gold standard mouse smoke exposure model to identify significant latent factors (context-specific co-expression modules) impacting pathophysiology. The machine learning transcriptomic signatures coupled to protein networks uncovered a reduction in network complexity and novel biological alterations in actin-associated gelsolin (GSN), which was transcriptionally linked to disease state. GSN was altered in airway epithelial cells in the mouse model and in human COPD. GSN was increased in plasma from COPD patients, and smoke exposure resulted in enhanced GSN release from airway cells from COPD patients. This method provides insights into rewiring of transcriptional networks that are associated with COPD pathogenesis and provide a novel analytical platform for other diseases.

Light field images super-resolution method based on hybrid low-dimensional spatial–angular interaction feature and linear complementation epipolar feature

Light-field (LF) cameras record 4D information about the intensity and angle of light, but limited by the resolution of the sensor, LF images require a trade-off between spatial and angular resolution, which results in generally low spatial resolution for LF images. We work on spatial super-resolution reconstruction (SR) of LF images, but due to the high dimensionality of 4D LF, it is difficult to process it directly, so we decompose 4D LF into low-dimensional sub-spaces. Since the 4D features of interest to different low-dimensional sub-spaces are different, we design a hybrid network structure—HSAESR, which processes multiple low-dimensional sub-spaces individually to adequately model the texture information in the 4D LF. HSAESR is designed to reconstruct Macro Pixel Images (MacPI) and Extremely Planar Images (EPI) respectively, and the final reconstruction results are weighted and fused. For MacPI, a CNN-based feature interaction module is designed using a novel asymptotic ’grid’ interaction approach to fully utilize the correlation between the 2D spatial dimension and the 2D angular dimension sub-spaces in 4D LF. For EPI, a Transformer-based epipolar feature linear complementation module is designed learning global and non-global multi-directional epipolar feature correlations, which learns feature linear correlation factor (α) and linear bias weight (β) through the correlation of non-global epipolar features, and corrects and complements the global epipolar correlation features through feature linear complementation. Extensive experiments are conducted on five datasets, the results show that HSAESR further improves the performance of LF images spatial SR, the reconstruction results are better than the comparison methods.

Machine learning based classification of excessive smartphone users via neuronal cue reactivity

Excessive Smartphone Use (ESU) poses a significant challenge in contemporary society, yet its recognition as a distinct disorder remains ambiguous. This study aims to address this gap by leveraging functional magnetic resonance imaging (fMRI) data and machine learning techniques to classify ESU and non-excessive smartphone users (n-ESU) based on their neural Cue-Reactivity (CR) signatures. By conducting a CR task and analyzing brain activation patterns, we identified spatial similarities between addictive smartphone use and established addictive disorders. Our approach involved employing Support Vector Machines (SVM) for classification, enhanced with feature selection methods such as Recursive Feature Elimination (RFE) and Model-based Selection and dimensionality reduction methods such as and Principal Component Analysis (PCA) to mitigate the challenges posed by limited dataset size and high dimensionality of fMRI data. The results demonstrate the effectiveness of our classification model, achieving accuracies of up to 79.9 %. Furthermore, we observed region-specific activations contributing significantly to classification accuracy, highlighting the potential biomarkers associated with ESU. External validation on longitudinal data revealed the necessity for larger training datasets to improve model generalizability. Additionally, feature selection techniques proved crucial for optimizing model performance, particularly in datasets with combined information from multiple sources. Our findings underscore the importance of incorporating more data to enhance model stability and generalizability, with implications for advancing the understanding and treatment of ESU and related disorders. Overall, our study demonstrates the promise of machine learning approaches in elucidating neural correlates of ESU and informing targeted interventions for affected individuals.

Effect of Social Dimension of Home on Academic Performance of Secondary School Students of District Buner

Students' academic performance has been highly valued and considered a major educational goal. This study is an attempt to explore the effect of the social dimension of home on the academic performance of District Buner secondary school students. The population of the current study was students of class 10th in District Buner. Cluster Random sampling technique was used to select the sample from the population. A standardized tool was used to measure the social dimension of the home of secondary school students. The data were analyzed quantitatively using mean, Standard Deviation, and Pearson correlation coefficient (‘r’) and results revealed that the children with a high social dimension of the home have higher academic performance. Level of Control is positively correlated with academic performance in the case of girls and negatively correlated in the case of boys. Permissiveness correlated negatively with academic performance for both girls and boys. Parents are recommended to create a home where everyone feels safe and loved. The government and media should also periodically conduct seminars and workshops to educate parents on improving social dimensions in the home environment that enhance children's overall development.

Fractal-based supervised approach for dimensionality reduction of hyperspectral images

Dimensionality reduction is one of the most challenging and crucial issues apart from data mining, security, and scalability, which have retained much traction due to the ever-growing need to analyze the large volumes of data generated daily. Fractal Dimension (FD) has been successfully used to characterize data sets and has found relevant applications in dimension reduction. This paper presents an application of the FD Reduction (FDR) Algorithm on geospatial hyperspectral data, examining its usefulness for data sets with a relatively high embedding dimension. We examine the algorithm at two levels. First is the conventional FDR approach (unsupervised) at the image level. Alternatively, we propose a pixel-level supervised approach for band reduction based on time-series complexity analysis. Techniques for determining an optimal intrinsic dimension for the dataset using these two techniques are examined. We also develop a parallel GPU-based implementation for the unsupervised image-level FDR algorithm, reducing the run-time by nearly 10 times. Furthermore, both approaches use a support vector machine classifier to compare the classification performance of the original and reduced image obtained.

A Hybrid Approach for Face Morphing Detection

Background: In biometrics, one of the most popular study topics is the detection of face morphing attacks. However, because present methods are unable to capture significant feature changes, they are unable to strike the correct balance between accuracy and complexity. Survey investigation and analysis have shown that the existing method of face morphing detection take a bit longer time to detect the image attack due to the high computation required by facial feature extraction approaches. Conversely, further study is needed to develop a model to enhance the computational time and accuracy of the current face morphing recognition methods. The paper developed a hybrid model for face morphing detection. The FERET database was created to aid in the evaluation and development of algorithms. Local Binary Pattern (LBP) was used as feature extraction algorithm and Residue Number System (RNS) was introduced to reduce the lengthy computational time of LBP during the extraction of images. The classification accuracy of 98% was achieved for the FERET database, while an accuracy of 96% was achieved for the FRGCv2 database. An average training time of 0.0532seconds was recorded for the FERET database, while an average training time of 0.0582seconds was achieved for the FRGCv2 database. The study concluded that the high dimensionality of LBP was well reduced and optimized by the RNS algorithm, which improved the performance of face morphing recognition.

Neutral Benchmarking of Survival Models in Health Sciences: Comparative Study of Classical and Machine Learning Techniques

Survival analysis plays a central role in diverse research fields, especially in health sciences. As an analytical tool, it can be used to help improve patients’ survival time, or at least, reduce the prospects of recurrence in cancer studies. However, approaches to the predictive performance of the current survival models mainly center on clinical data along with the classical survival methods. For censored “omics” data, the performance of survival models has not been thoroughly studied, either often due to their high dimensionality issues or reliance on binarizing the survival time for classification analysis. We aim to present a neural benchmark approach that analyzes and compares a broad range of classical and state-of-the-art machine learning survival models for “omics” and clinical datasets. All the methods considered in our study are evaluated using predictability as a performance measure. The study is systematically designed to make 36 comparisons (9 methods over 4 datasets, i.e., 2 clinical and 2 omics), and shows that, in practice, predictability of survival models does vary across real-world datasets, model choice, as well as the evaluation metric. From our results, we emphasize that performance criteria can play a key role in a balanced assessment of diverse survival models. Moreover, the Multitask Logistic Regression (MTLR) showed remarkable predictability for almost all the datasets. We believe this outstanding performance presents a unique opportunity for a wider use of MTLR for survival risk factors. For translational clinicians and scientists, we hope our findings provide practical guidance for benchmark studies of survival models, as well as highlight potential areas of research interest.

A Representation-Learning-Based Graph and Generative Network for Hyperspectral Small Target Detection

Hyperspectral small target detection (HSTD) is a promising pixel-level detection task. However, due to the low contrast and imbalanced number between the target and the background spatially and the high dimensions spectrally, it is a challenging one. To address these issues, this work proposes a representation-learning-based graph and generative network for hyperspectral small target detection. The model builds a fusion network through frequency representation for HSTD, where the novel architecture incorporates irregular topological data and spatial–spectral features to improve its representation ability. Firstly, a Graph Convolutional Network (GCN) module better models the non-local topological relationship between samples to represent the hyperspectral scene’s underlying data structure. The mini-batch-training pattern of the GCN decreases the high computational cost of building an adjacency matrix for high-dimensional data sets. In parallel, the generative model enhances the differentiation reconstruction and the deep feature representation ability with respect to the target spectral signature. Finally, a fusion module compensates for the extracted different types of HS features and integrates their complementary merits for hyperspectral data interpretation while increasing the detection and background suppression capabilities. The performance of the proposed approach is evaluated using the average scores of AUCD,F, AUCF,τ, AUCBS, and AUCSNPR. The corresponding values are 0.99660, 0.00078, 0.99587, and 333.629, respectively. These results demonstrate the accuracy of the model in different evaluation metrics, with AUCD,F achieving the highest score, indicating strong detection performance across varying thresholds. Experiments on different hyperspectral data sets demonstrate the advantages of the proposed architecture.

ScGAA: a general gated axial-attention model for accurate cell-type annotation of single-cell RNA-seq data

Single-cell RNA sequencing (scRNA-seq) is a key technology for investigating cell development and analysing cell diversity across various diseases. However, the high dimensionality and extreme sparsity of scRNA-seq data pose great challenges for accurate cell type annotation. To address this, we developed a new cell-type annotation model called scGAA (general gated axial-attention model for accurate cell-type annotation of scRNA-seq). Based on the transformer framework, the model decomposes the traditional self-attention mechanism into horizontal and vertical attention, considerably improving computational efficiency. This axial attention mechanism can process high-dimensional data more efficiently while maintaining reasonable model complexity. Additionally, the gated unit was integrated into the model to enhance the capture of relationships between genes, which is crucial for achieving an accurate cell type annotation. The results revealed that our improved transformer model is a promising tool for practical applications. This theoretical innovation increased the model performance and provided new insights into analytical tools for scRNA-seq data.

Stealthy data integrity attack identification in smart grid networks utilizing deep denoising autoencoder

Smart grids arose as the largest cyber-physical systems with the integration of sophisticated control, computing, and state-of-the-art communications. Like all cyber-physical systems, the smart grids are vulnerable to malicious cyber assaults due to their enormous dependency on communication networks. Various machine learning-based schemes are being investigated in the industry and academia to develop robust defense mechanisms to counter cyber assaults. However, the curse of high dimensionality, which increases with the escalating evolution of an electric power system, infringes upon the efficiency of machine learning models employed to detect such assaults. To this end, this paper proposes a deep denoising autoencoder (DAE)-based framework for dimensionality reduction that learns salient feature representation for high-dimensional, multi-variant smart grid measurement data collected from the smart grids. The latent space apprehended by DAE is then fed to binary support vector machine (SVM) to determine the assaulted data. Various standard IEEE test cases are employed in simulations. The results show that the proposed scheme learns more robust features that reveal the nonlinear properties exhibited in the smart grid measurements, further leading to improved detection accuracy of the classifier as compared to existing approaches.

Feedback-based quantum algorithms for ground state preparation

The ground state properties of quantum many-body systems are a subject of interest across chemistry, materials science, and physics. Thus, algorithms for finding ground states can have broad impacts. Variational quantum algorithms are one class of ground state algorithms that has received significant attention in recent years. These algorithms utilize a hybrid quantum-classical computing framework to prepare ground states on quantum computers. However, this requires solving a classical optimization problem that can become prohibitively expensive in high dimensions. Here, we develop formulations of feedback-based quantum algorithms for ground state preparation that can be used to address this challenge for two broad classes of Hamiltonians: Fermi-Hubbard Hamiltonians, and molecular Hamiltonians represented in second quantization. Feedback-based quantum algorithms are optimization-free; in place of classical optimization, quantum circuit parameters are set according to a deterministic feedback law derived from quantum Lyapunov control principles. This feedback law guarantees a monotonic improvement in solution quality with respect to the depth of the quantum circuit. A variety of numerical illustrations are provided that analyze the convergence and robustness of feedback-based quantum algorithms for these problem classes. Published by the American Physical Society 2024

Factors Controlling Differences in Morphology and Fractal Characteristics of Organic Pores of Longmaxi Shale in Southern Sichuan Basin, China

Shale gas is a prospective cleaner energy resource and the exploration and development of shale gas has made breakthroughs in many countries. Structure deformation is one of the main controlling factors of shale gas accumulation and enrichment in complex tectonic areas in southern China. In order to estimate the shale gas capacity of structurally deformed shale reservoirs, it is necessary to understand the systematic evolution of organic pores in the process of structural deformation. In particular, as the main storage space of high-over-mature marine shale reservoirs, the organic matter pore system directly affects the occurrence and migration of shale gas; however, there is a lack of systematic research on the fractal characteristics and deformation mechanism of organic pores under the background of different tectonic stresses. Therefore, to clarify the above issues, modular automated processing system (MAPS) scanning, low-pressure gas adsorption, quantitative evaluation of minerals by scanning (QEMSCAN), and focused ion beam scanning electron microscopy (FIB-SEM) were performed and interpreted with fractal and morphology analyses to investigate the deformation mechanisms and structure of organic pores from different tectonic units in Silurian Longmaxi shale. Results showed that in stress concentration areas such as around veins or high-angle fractures, the organic pore length-width ratio and the fractal dimension are higher, indicating that the pore is more obviously modified by stress. Under different tectonic backgrounds, the shale reservoir in Weiyuan suffered severe denudation and stronger tectonic compression during burial, which means that the organic pores are dominated by long strip pores and slit-shaped pores with high fractal dimension, while the pressure coefficient in Luzhou is high and the structural compression is weak, resulting in suborbicular pores and ink bottle pores with low fractal dimension. The porosity and permeability of different forms of organic pores are also obviously different; the connectivity of honeycomb pores with the smallest fractal dimension is the worst, that of suborbicular organic pores is medium, and that of long strip organic pores with the highest fractal dimension is the best. This study provides more mechanism discussion and case analysis for the microscopic heterogeneity of organic pores in shale reservoirs and also provides a new analysis perspective for the mechanism of shale gas productivity differences in different stress–strain environments.

Hyperdimensional computing: A fast, robust, and interpretable paradigm for biological data.

Advances in bioinformatics are primarily due to new algorithms for processing diverse biological data sources. While sophisticated alignment algorithms have been pivotal in analyzing biological sequences, deep learning has substantially transformed bioinformatics, addressing sequence, structure, and functional analyses. However, these methods are incredibly data-hungry, compute-intensive, and hard to interpret. Hyperdimensional computing (HDC) has recently emerged as an exciting alternative. The key idea is that random vectors of high dimensionality can represent concepts such as sequence identity or phylogeny. These vectors can then be combined using simple operators for learning, reasoning, or querying by exploiting the peculiar properties of high-dimensional spaces. Our work reviews and explores HDC's potential for bioinformatics, emphasizing its efficiency, interpretability, and adeptness in handling multimodal and structured data. HDC holds great potential for various omics data searching, biosignal analysis, and health applications.

PGNN-Net: Parallel Graph Neural Networks for Hyperspectral Image Classification Using Multiple Spatial-Spectral Features

Hyperspectral image (HSI) shows great potential for application in remote sensing due to its rich spectral information and fine spatial resolution. However, the high dimensionality, nonlinearity, and complex relationship between spectral and spatial features of HSI pose challenges to its accurate classification. Traditional convolutional neural network (CNN)-based methods suffer from detail loss in feature extraction; Transformer-based methods rely too much on the quantity and quality of HSI; and graph neural network (GNN)-based methods provide a new impetus for HSI classification by virtue of their excellent ability to handle irregular data. To address these challenges and take advantage of GNN, we propose a network of parallel GNNs called PGNN-Net. The network first extracts the key spatial-spectral features of HSI using principal component analysis, followed by preprocessing to obtain two primary features and a normalized adjacency matrix. Then, a parallel architecture is constructed using improved GCN and ChebNet to extract local and global spatial-spectral features, respectively. Finally, the discriminative features obtained through the fusion strategy are input into the classifier to obtain the classification results. In addition, to alleviate the over-fitting problem, the label smoothing technique is embedded in the cross-entropy loss function. The experimental results show that the average overall accuracy obtained by our method on Indian Pines, Kennedy Space Center, Pavia University Scene, and Botswana reaches 97.35%, 99.40%, 99.64%, and 98.46%, respectively, which are better compared to some state-of-the-art methods.

Identifying Proteomic Prognostic Markers for Alzheimer's Disease with Survival Machine Learning: the Framingham Heart Study.

Protein abundance levels, sensitive to both physiological changes and external interventions, are useful for assessing the Alzheimer's disease (AD) risk and treatment efficacy. However, identifying proteomic prognostic markers for AD is challenging by their high dimensionality and inherent correlations. Our study analyzed 1128 plasma proteins, measured by the SOMAscan platform, from 858 participants 55 years and older (mean age 63 years, 52.9% women) of the Framingham Heart Study (FHS) Offspring cohort. We conducted regression analysis and machine learning models, including LASSO-based Cox proportional hazard regression model (LASSO) and generalized boosted regression model (GBM), to identify protein prognostic markers. These markers were used to construct a weighted proteomic composite score, the AD prediction performance of which was assessed using time-dependent area under the curve (AUC). The association between the composite score and memory domain was examined in 339 (of the 858) participants with available memory scores, and in an independent group of 430 participants younger than 55 years (mean age 46, 56.7% women). Over a mean follow-up of 20 years, 132 (15.4%) participants developed AD. After adjusting for baseline age, sex, education, and APOE ε4+ status, regression models identified 309 proteins (P ≤ 0.2). After applying machine learning methods, nine of these proteins were selected to develop a composite score. This score improved AD prediction beyond the factors of age, sex, education, and APOE ε4+ status across 15 to 25 years of follow-up, achieving its peak AUC of 0.84 in the LASSO model at the 22-year follow-up. It also showed a consistent negative association with memory scores in 339 participants (beta = -0.061, P = 0.046), 430 independent participants (beta = -0.060, P = 0.018), and the pooled 769 samples (beta = -0.058, P = 0.003). These findings highlight the utility of proteomic markers in improving AD prediction and emphasize the complex pathology of AD. The composite score may aid early AD detection and efficacy monitoring, warranting further validation in diverse populations.

ScTCA: a hybrid Transformer-CNN architecture for imputation and denoising of scDNA-seq data.

Single-cell DNA sequencing (scDNA-seq) has been widely used to unmask tumor copy number alterations (CNAs) at single-cell resolution. Despite that arm-level CNAs can be accurately detected from single-cell read counts, it is difficult to precisely identify focal CNAs as the read counts are featured with high dimensionality, high sparsity and low signal-to-noise ratio. This gives rise to a desperate demand for reconstructing high-quality scDNA-seq data. We develop a new method called scTCA for imputation and denoising of single-cell read counts, thus aiding in downstream analysis of both arm-level and focal CNAs. scTCA employs hybrid Transformer-CNN architectures to identify local and non-local correlations between genes for precise recovery of the read counts. Unlike conventional Transformers, the Transformer block in scTCA is a two-stage attention module containing a stepwise self-attention layer and a window Transformer, and can efficiently deal with the high-dimensional read counts data. We showcase the superior performance of scTCA through comparison with the state-of-the-arts on both synthetic and real datasets. The results indicate it is highly effective in imputation and denoising of scDNA-seq data.

Multi-level multi-view network based on structural contrastive learning for scRNA-seq data clustering.

Clustering plays a crucial role in analyzing scRNA-seq data and has been widely used in studying cellular distribution over the past few years. However, the high dimensionality and complexity of scRNA-seq data pose significant challenges to achieving accurate clustering from a singular perspective. To address these challenges, we propose a novel approach, called multi-level multi-view network based on structural consistency contrastive learning (scMMN), for scRNA-seq data clustering. Firstly, the proposed method constructs shallow views through the $k$-nearest neighbor ($k$NN) and diffusion mapping (DM) algorithms, and then deep views are generated by utilizing the graph Laplacian filters. These deep multi-view data serve as the input for representation learning. To improve the clustering performance of scRNA-seq data, contrastive learning is introduced to enhance the discrimination ability of our network. Specifically, we construct a group contrastive loss for representation features and a structural consistency contrastive loss for structural relationships. Extensive experiments on eight real scRNA-seq datasets show that the proposed method outperforms other state-of-the-art methods in scRNA-seq data clustering tasks. Our source code has already been available at https://github.com/szq0816/scMMN.

Graph contrastive learning as a versatile foundation for advanced scRNA-seq data analysis.

Single-cell RNA sequencing (scRNA-seq) offers unprecedented insights into transcriptome-wide gene expression at the single-cell level. Cell clustering has been long established in the analysis of scRNA-seq data to identify the groups of cells with similar expression profiles. However, cell clustering is technically challenging, as raw scRNA-seq data have various analytical issues, including high dimensionality and dropout values. Existing research has developed deep learning models, such as graph machine learning models and contrastive learning-based models, for cell clustering using scRNA-seq data and has summarized the unsupervised learning of cell clustering into a human-interpretable format. While advances in cell clustering have been profound, we are no closer to finding a simple yet effective framework for learning high-quality representations necessary for robust clustering. In this study, we propose scSimGCL, a novel framework based on the graph contrastive learning paradigm for self-supervised pretraining of graph neural networks. This framework facilitates the generation of high-quality representations crucial for cell clustering. Our scSimGCL incorporates cell-cell graph structure and contrastive learning to enhance the performance of cell clustering. Extensive experimental results on simulated and real scRNA-seq datasets suggest the superiority of the proposed scSimGCL. Moreover, clustering assignment analysis confirms the general applicability of scSimGCL, including state-of-the-art clustering algorithms. Further, ablation study and hyperparameter analysis suggest the efficacy of our network architecture with the robustness of decisions in the self-supervised learning setting. The proposed scSimGCL can serve as a robust framework for practitioners developing tools for cell clustering. The source code of scSimGCL is publicly available at https://github.com/zhangzh1328/scSimGCL.