Feature Clustering Research Articles

Non-negative consistency affinity graph learning for unsupervised feature selection and clustering

Feature selection plays a crucial role in data mining and pattern recognition tasks. This paper proposes an efficient method for robust unsupervised feature selection, called joint local preserving and low-rank representation with nonnegative and symmetric constraint (JLPLRNS). This approach utilizes an indicator matrix instead of the row sparsity of the projection matrix to directly select some significant features from the original data and adaptively preserves the local geometric structure of original data into the low-dimensional embedded feature subspace via learned indicator matrix. By simultaneously imposing nonnegative symmetric and low-rank constraints on the representation coefficient matrix, it cannot only make this matrix discriminative, sparse and weight consistency for each pair of data, but also uncover the global structure of original data. These effectively will improve clustering performance. An algorithm based on the augmented Lagrange multiplier method with an alternating direction strategy is designed to resolve this model. Experimental results on various real datasets show that the proposed method can effectively identify some important features in data and outperforms many state-of-the-art unsupervised feature selection methods in terms of clustering performance.

Epidemiological characteristics and spatial aggregation of acute myocardial infarction in Shandong Province

Objective: To understand the characteristics and trends of acute myocardial infarction (AMI) in Shandong Province and to provide evidence for formulating prevention and control strategies. Methods: Data were derived from the AMI incidence reports of Shandong Province's Chronic Disease Surveillance Information Management System in 2012-2021. The crude and standardized incidence rates were used as indicators to describe the incidence level of AMI. Joinpoint regression analysis was used to analyze the trends in the incidence and age of onset over the years. The contribution of population aging to the increase in AMI incidence was assessed using the rate difference decomposition method. The incidence of AMI in each district (county) in Shandong Province was visualized using ArcGIS 10.8 software, and global and local spatial autocorrelation analysis was performed using DeoDa 1.12 software. Results: From 2012 to 2021, 198 233 cases of AMI were reported from 19 provincial monitoring sites in Shandong Province, of which 53.13% were males and 97.12% were ≥45 years old. The reported crude incidence increased from 90.12 per 100 000 in 2012 to 176.54 per 100 000 in 2021, with an average annual increase of 7.01% (Z=7.35, P<0.001). There was no significant upward trend in standardized incidence (Z=1.64, P=0.140), but the standardized incidence of male residents showed an increasing trend (Z=2.76, P=0.028). Before 2014, the reported crude incidence of males was similar to that of females, but after 2014, the reported crude incidence of males was continuously higher than that of females. However, males' standardized incidence was higher than females in all years. Both crude and standardized incidence rates were higher in rural residents than in urban areas. The median onset of AMI increased from 71.6 years old in 2012 to 73.5 years old in 2021. The median age of onset in males was lower than that in females in all years, and in most years, the median age of onset in urban residents was lower than that in rural residents. The incidence of AMI in males showed a trend in younger age groups. According to the seasonal decomposition, the incidence peak of AMI was in January, and the trough was in September. The contribution of aging population to the increase in crude incidence of AMI increased from 8.63% in 2013 to 52.58% in 2021. The global spatial autocorrelation analysis showed that the incidence of AMI presented an obvious spatial clustering distribution. Local spatial autocorrelation analysis found that the high-incidence areas (counties) were mainly concentrated in Liaocheng City and Dezhou City in the northwest region of Shandong Province and Heze City in the southwest. Conclusions: The incidence of AMI among residents in Shandong Province was rising, with spatial clustering and seasonal clustering characteristics. People aged 45 years and older, male residents, and rural residents were at high risk of developing AMI. There was a certain trend of younger age at onset among men. Targeted prevention and control measures should be taken for high-incidence seasons, high-risk groups, and high-incidence clustering areas in northwestern Shandong Province.

Design of short-term load forecasting method considering user behavior

In order to ensure optimize the short-term power load forecasting effect, load forecasting method design taking into account the user behavior is studied. Mining user power consumption behavior data from power user power consumption information acquisition system, and suppressing power consumption behavior data noise through empirical mode decomposition algorithm; Input the denoised user behavior data into the improved Relief algorithm, extract the user power consumption behavior characteristics, eliminate redundant features through correlation analysis, and cluster the user behavior characteristics using K-means clustering algorithm to reduce the dimension of the feature data and the complexity of the features, then input the clustered features into the online extreme learning machine (OS-ELM), load forecasting is realized through online learning. Experiments show that this method can accurately predict the power consumption of each type of users in different time periods, and its prediction results have a high interval coverage.

Unsupervised clustering analysis based on multidimensional features reveals distinct clinical characteristics and associated factors of different phenotypes in patients with chronic rhinosinusitis with nasal polyp

Objective: To utilize routinely available clinical parameters to uncover the clinical features of different clusters in patients with chronic rhinosinusitis with nasal polyp (CRSwNP) through unsupervised clustering analysis. Methods: The clinical data from 155 CRSwNP patients undergoing nasal endoscopic surgery at Renmin Hospital of Wuhan University from 2021 to 2023 were prospectively collected, including 112 males and 43 females, aged from 7 to 87 years. Unsupervised clustering analysis was conducted using various clinical parameters, including age, gender, smoking and drinking history, local eosinophil (EOS) and neutrophil (NEU) counts, comorbid allergic rhinitis (AR), comorbid asthma, recurrence status, serum-specific IgE, total IgE, cytokine levels, peripheral blood EOS count and percentage, Lund-Mackay CT score, the ratio of CT scores for the maxillary sinus and ethmoid sinus (E/M ratio), visual analogue scale (VAS) score, Lund-Kennedy endoscopic score, and other common clinical indicators to elucidate the clinical characteristics of each cluster. Statistical analysis was conducted using GraphPad Prism 9.5 software. Results: Hierarchical clustering analysis identified four main clusters (Cluster A1-A4), showcasing distinct characteristics such as mild nasal polyps with higher peripheral blood cytokines levels, nasal polyps accompanied by allergies and asthma, a subtype of nasal polyps with high recurrence rates dominated by neutrophils, and nasal polyps with high eosinophil levels. Further subset clustering revealed two clusters of mild polyps (Cluster B1-B2) featuring high cytokine expression and comorbid AR; and two clusters of severe polyps (Cluster B3-B4) presented with severe symptoms, high Lund-Mackay CT score, and high Lund-Kennedy endoscopic score. Variations between Cluster B3 and B4 included symptom complexity, the degree of eosinophil infiltration, and the probability of comorbid asthma. Further clustering analysis for eosinophilic nasal polyps revealed a cluster characterized by highly neutrophilic infiltration and recurrent nasal polyps. The comprehensive analysis of multi-index correlations demonstrated valuable insights into the relationships between common clinical parameters of nasal polyps, providing valuable information for a deeper understanding of the pathogenesis of CRSwNP. Conclusion: The clustering analysis in this study categorizes CRSwNP patients into different clusters based on clinical features and disease outcomes, providing a new perspective for more precise clinical treatment strategies.

Does consensus contour improve robustness and accuracy in 18F-FDG PET radiomic features?

PurposeThe purpose of our study is to validate the robustness and accuracy of consensus contour in 2-deoxy-2-[18\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{18}$$\\end{document}F]fluoro-D-glucose (18\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{18}$$\\end{document}F-FDG) PET radiomic features.Methods225 nasopharyngeal carcinoma (NPC) and 13 extended cardio-torso (XCAT) simulated data were enrolled. All segmentation were performed with four segmentation methods under two different initial masks, respectively. Consensus contour (ConSeg) was then developed using the majority vote rule. 107 radiomic features were extracted by Pyradiomics based on segmentation and the intraclass correlation coefficient (ICC) was calculated for each feature between masks or among segmentation, respectively. In XCAT ICC between segmentation and simulated ground truth were also calculated to access the accuracy.ResultsICC varied with the dataset, segmentation method, initial mask and feature type. ConSeg presented higher ICC for radiomic features in robustness tests and similar ICC in accuracy tests, compared with the average of four segmentation results. Higher ICC were also generally observed in irregular initial masks compared with rectangular masks in both robustness and accuracy tests. Furthermore, 19 features (17.76%) had ICC ≥ 0.75 in both robustness and accuracy tests for any of the segmentation methods or initial masks. The dataset was observed to have a large impact on the correlation relationships between radiomic features, but not the segmentation method or initial mask.ConclusionsThe consensus contour combined with irregular initial mask could improve the robustness and accuracy in radiomic analysis to some extent. The correlation relationships between radiomic features and feature clusters largely depended on the dataset, but not segmentation method or initial mask.

Assessment of ecosystem health and driving forces in response to landscape pattern dynamics: the Shibing Karst world natural heritage site case study

The Shibing Karst constitutes a pivotal component of the "South China Karst," and its ecosystem health integrity crucially influences the Outstanding Universal Value (OUV) of the corresponding Natural World Heritage (NWH). Consequently, robust ecosystem health assessment (EHA) is imperative for the judicious conservation and management of this heritage, as well as for the sustainable progression of the region. This research assessed the health of the Shibing Karst ecosystem from 2004 to 2020 by employing changes in landscape patterns through the Vigor-Organization-Resilience-Ecosystem Services (VORS) model. Spatial autocorrelation was employed to discern the spatial coherence and evolutionary patterns of ecosystem health, whereas a geo-detector ascertained the pivotal determinants impacting regional ecosystem vitality. The findings revealed that: (1) The landscape patterns distribution in the study area exhibited considerable constancy, primarily comprising forest land, with a rising trajectory in construction land and water, juxtaposed with a recession in shrubland, grassland, paddy land, and dryland expanses. (2) From 2004 to 2020, the ecosystem of the study area maintained its health and remained stable, with mean values of 0.8303, 0.7689, 0.6976, and 0.7824, respectively, showing an evolutionary trend of an initial downtrend trend followed by an upswing, with 2016 marking a pivotal juncture. (3) Spatial clustering analysis highlighted significant clustering characteristics of ecosystem health, with a nominal decrease in the Global Moran's I index from 0.666 to 0.665, which is indicative of a subtle decrease in clustering over time. High-high clustering areas were predominantly located within the World Heritage Site (WHS), while low-low clustering areas were mainly distributed in the southeastern part of buffer zone. (4) Land use and cover change (LUCC) and Ecosystem Services (ESs) were identified as the primary indexes of EHA, with Ecosystem Resilience (ER), Ecosystem Vigor (EV), and Ecosystem Organization (EO) exerting relatively mild influences. This study provides a scientific framework for policymakers in local governance to devise strategies for ecosystem conservation and management, enhances the analytical perspective on the integrity and conservation of Karst Natural World Heritage (KNWH).

Dynamics and features of transmission clusters of HIV-1 subtypes in the state of São Paulo, Brazil.

Molecular epidemiology techniques allow us to track the HIV-1 transmission dynamics. Herein, we combined genetic, clinical and epidemiological data collected during routine clinical treatment to evaluate the dynamics and characteristics of transmission clusters of the most prevalent HIV-1 subtypes in the state of São Paulo, Brazil. This was a cross-sectional study conducted with 2,518 persons living with HIV (PLWH) from 53 cities in São Paulo state between Jan 2004 to Feb 2015. The phylogenetic tree of protease/reverse transcriptase (PR/RT) regions was reconstructed by PhyML and ClusterPicker used to infer the transmission clusters based on Shimodaira-Hasegawa (SH) greater than 90% (phylogenetic support) and genetic distance less than 6%. Of a total of 2,518 sequences, 2,260 were pure subtypes at the PR/RT region, being B (88%), F1 (8.1%), and C (4%). About 21.2% were naïve with a transmitted drug resistance (TDR) rate of 11.8%. A total of 414 (18.3%) of the sequences clustered. These clusters were less evident in subtype B (17.7%) and F1 (15.1%) than in subtype C (40.2%). Clustered sequences were from PLWH at least 5 years younger than non-clustered among subtypes B (p < 0.001) and C (p = 0.037). Men who have sex with men (MSM) predominated the cluster in subtype B (51%), C (85.7%), and F1 (63.6%; p < 0.05). The TDR rate in clustered patients was 15.4, 13.6, and 3.1% for subtypes B, F1, and C, respectively. Most of the infections in subtypes B (80%), C (64%), and F1 (59%) occurred within the state of São Paulo. The metropolitan area of São Paulo presented a high level of endogenous clustering for subtypes B and C. The São Paulo city had 46% endogenous clusters of subtype C. Our findings showed that MSM, antiretroviral therapy in Treatment-Naive (ART-naïve) patients, and HIV1-C, played an important role in the HIV epidemic in the São Paulo state. Further studies in transmission clusters are needed to guide the prevention intervention.

Carborane-Cluster-Wrapped Copper Cluster with Cyclodextrin-like Cavities for Chiral Recognition.

Chiral atomically precise metal clusters, known for their remarkable chiroptical properties, hold great potential for applications in chirality recognition. However, advancements in this field have been constrained by the limited exploration of host-guest chemistry, involving metal clusters. This study reports the synthesis of a chiral Cu16(C2B10H10S2)8 (denoted as Cu16@CB8, where C2B10H12S2H2 = 9,12-(HS)2-1,2-closo-carborane) cluster by an achiral carboranylthiolate ligand. The chiral R-/S-Cu16@CB8 cluster features chiral cavities reminiscent of cyclodextrins, which are surrounded by carborane clusters, yet they crystallize in a racemate. These cyclodextrin-like cavities demonstrated the specific recognition of amino acids, as indicated by the responsive output of circular dichroism and circularly polarized luminescence signals of Cu16 moieties of the Cu16@CB8 cluster. Notably, a quantitative chiroptical analysis of amino acids in a short time and a concomitant deracemization of Cu16@CB8 were achieved. Density functional tight-binding molecular dynamics simulation and noncovalent interaction analysis further unraveled the great importance of the cavities and binding sites for chiral recognition. Dipeptide, tripeptide, and polypeptide containing the corresponding amino acids (Cys, Arg, or His residues) display the same chiral recognition, showing the generality of this approach. The functional synergy of dual clusters, comprising carborane and metal clusters, is for the first time demonstrated in the Cu16@CB8 cluster, resulting in the valuable quantification of the enantiomeric excess (ee) value of amino acids. This work opens a new avenue for chirality sensors based on chiral metal clusters with unique chiroptical properties and inspires the development of carborane clusters in host-guest chemistry.

Investigation of geographic disparities of diabetes-related hospitalizations in Florida using flexible spatial scan statistics: An ecological study.

Hospitalizations due to diabetes complications are potentially preventable with effective management of the condition in the outpatient setting. Diabetes-related hospitalization (DRH) rates can provide valuable information about access, utilization, and efficacy of healthcare services. However, little is known about the local geographic distribution of DRH rates in Florida. Therefore, the objectives of this study were to investigate the geographic distribution of DRH rates at the ZIP code tabulation area (ZCTA) level in Florida, identify significant local clusters of high hospitalization rates, and describe characteristics of ZCTAs within the observed spatial clusters. Hospital discharge data from 2016 to 2019 were obtained from the Florida Agency for Health Care Administration through a Data Use Agreement with the Florida Department of Health. Raw and spatial empirical Bayes smoothed DRH rates were computed at the ZCTA level. High-rate DRH clusters were identified using Tango's flexible spatial scan statistic. Choropleth maps were used to display smoothed DRH rates and significant high-rate spatial clusters. Demographic, socioeconomic, and healthcare-related characteristics of cluster and non-cluster ZCTAs were compared using the Wilcoxon rank sum test for continuous variables and Chi-square test for categorical variables. There was a total of 554,133 diabetes-related hospitalizations during the study period. The statewide DRH rate was 8.5 per 1,000 person-years, but smoothed rates at the ZCTA level ranged from 0 to 101.9. A total of 24 significant high-rate spatial clusters were identified. High-rate clusters had a higher percentage of rural ZCTAs (60.9%) than non-cluster ZCTAs (41.8%). The median percent of non-Hispanic Black residents was significantly (p < 0.0001) higher in cluster ZCTAs than in non-cluster ZCTAs. Populations of cluster ZCTAs also had significantly (p < 0.0001) lower median income and educational attainment, and higher levels of unemployment and poverty compared to the rest of the state. In addition, median percent of the population with health insurance coverage and number of primary care physicians per capita were significantly (p < 0.0001) lower in cluster ZCTAs than in non-cluster ZCTAs. This study identified geographic disparities of DRH rates at the ZCTA level in Florida. The identification of high-rate DRH clusters provides useful information to guide resource allocation such that communities with the highest burdens are prioritized to reduce the observed disparities. Future research will investigate determinants of hospitalization rates to inform public health planning, resource allocation and interventions.

Regional differences, dynamic evolution, and obstacle factors of cultivated land ecological security in China

Under the influence of geopolitical conflicts, extreme weather, and other factors, global food security has been severely impacted. Cultivated land ecological security in China is vital for guaranteeing local food security and sustainable development. Therefore, using provincial data from 2003 to 2020, this research establishes an index system based on the pressure-state-response (PSR) model and assesses its cultivated land ecological security via the entropy weight method. A series of measurements such as Dagum's Gini coefficient and spatial Markov chain are then adopted to reveal changes in regional differences and spatial-temporal characteristics of cultivated land ecological security in China. The findings indicate that this ecological security has risen from insecure to critically secure, but the overall gap has gradually widened, mainly stemming from interregional differences. Cultivated land ecological security exhibits positive spatial clustering characteristics and the phenomenon of club convergence. The amount of water resources, the degree of soil erosion control, and the strength of local financial expenditure on agriculture are the main factors hindering its cultivated land ecological security, and the threats mainly come from the response criterion layer. Therefore, it is essential to strengthen measures for cropland ecological protection, to formulate targeted strategies by combining the regional characteristics, to narrow the intra-regional gap, and to bring into play the spatial radiation effect of high-level provinces.

Quantifying the features of earthquake clusters in north-Central Iran, based on nearest-neighbor distances and network analysis

The statistical features of earthquake clusters in North-Central Iran (Tehran Region) are investigated, with the aim of quantitatively characterizing the properties of earthquake triggering and allow exploring their possible relations with the tectonic setting of the study area.The nearest-neighbor approach is used for the identification of the earthquake clusters in the space-time-energy domain. This approach permits for a data-driven identification of clusters so that, within multi-event clusters, the features of secondary and higher orders dependent events can be explored. The study is based on a revised dataset that is extracted from the catalog compiled by the Iranian Seismological Center (IRSC) for the period of 1996–2022. In order to exclude the effect of non-tectonic events, which turn out quite numerous within the study region, explosions within quarry-rich areas are removed; the identification of non-tectonic events is performed by considering the normalized ratios of daytime to nighttime events in an iterative removal procedure. According to preliminary analysis of the resulting catalog, an area is selected, within which a satisfactory completeness level is assessed for events with magnitude >2.0. Robust values of the scaling parameters, namely the b-value and the fractal dimension of epicenters, are also computed and are used to calculate the nearest-neighbor distances and to identify the earthquake clusters.The nearest-neighbor method also permits to investigate the internal structure of earthquake sequences, and to differentiate the spatial properties of seismicity according to the different topological features of the clusters structure. The obtained results allow us identifying two macro-areas, approximately separated by the 52°E meridian, which are characterized by different clustering features, namely: high complexity indexes, indicating simple (burst-like) structure of clusters, to the East; low complexity index, corresponding to complex multi-level (swarm-like) structure of clusters, to the West. The complexity measures, borrowed from network theory (i.e. the Closeness and Outdegree Centralization indexes), consistently capture the complexity of the identified clusters, and confirm that the cluster structures have distinct preferred geographic locations. The territorial heterogeneity of the examined clustering properties can be related with the spatial variability of tectonic, structural and geophysical features of the Alborz region, in good agreement with findings from the Alps-Dinarides junction (Northeastern Italy), a region also characterized by a contractional structural setting, mainly including reverse and strike-slip faulting systems, and by moderate to high seismic activity.

CORE-ReID: Comprehensive Optimization and Refinement through Ensemble Fusion in Domain Adaptation for Person Re-Identification

This study introduces a novel framework, “Comprehensive Optimization and Refinement through Ensemble Fusion in Domain Adaptation for Person Re-identification (CORE-ReID)”, to address an Unsupervised Domain Adaptation (UDA) for Person Re-identification (ReID). The framework utilizes CycleGAN to generate diverse data that harmonize differences in image characteristics from different camera sources in the pre-training stage. In the fine-tuning stage, based on a pair of teacher–student networks, the framework integrates multi-view features for multi-level clustering to derive diverse pseudo-labels. A learnable Ensemble Fusion component that focuses on fine-grained local information within global features is introduced to enhance learning comprehensiveness and avoid ambiguity associated with multiple pseudo-labels. Experimental results on three common UDAs in Person ReID demonstrated significant performance gains over state-of-the-art approaches. Additional enhancements, such as Efficient Channel Attention Block and Bidirectional Mean Feature Normalization mitigate deviation effects and the adaptive fusion of global and local features using the ResNet-based model, further strengthening the framework. The proposed framework ensures clarity in fusion features, avoids ambiguity, and achieves high accuracy in terms of Mean Average Precision, Top-1, Top-5, and Top-10, positioning it as an advanced and effective solution for UDA in Person ReID.

Particle scale modeling of heat transfer between mesoscale clusters and surfaces in fluidized beds

The distribution of heat absorption and the mechanisms of conductive, convective, and radiative heat transfers between cold particle clusters (Tp = 300 K) and high-temperature surfaces (Tw = 1173 K) in fluidized beds were investigated through numerical simulations using the MFiX-DEM approach. The influence of cluster characteristics (εc and Dc), restitution coefficients (ep and ew), and particle properties (dp, ρp, kp, and cp) on heat absorption and various heat transfers was examined. The simulations indicate that convection predominantly governed the heat transfer behavior during the process of mesoscale clusters impacting the surfaces. A significant increase in heat transfer can be attained by reducing the cluster volume fraction, cluster size, particle diameter, and particle density. Besides, decreasing the restitution coefficients allowed for a relatively small enhancement of heat absorption. The maximum difference in ep was 2.11 kW/kg, while the maximum difference in ew was 0.99 kW/kg. The specific heat capacity and thermal conductivity of the particle had little influence on the distribution of heat absorption and heat transfer with a maximum relative error of up to 5.8 %. Finally, the mathematical correlation during surface-to-cluster heat transfer was established. This work introduces a novel application of the MFiX-DEM approach to analyze surface-to-cluster heat transfer at the particle scale, providing valuable insights for optimizing fluidized bed systems.

1DCAE-TSSAMC: Two-Stage Multi-Dimensional Spatial Features Based Multi-View Deep Clustering for Time Series Data

At present, as a research hotspot for time series data (TSD), the deep clustering analysis of TSD has huge research value and practical significance. However, there still exist the following three problems: (1) For deep clustering based on joint optimization, the inevitably mutual interference existing between deep feature representation learning progress and clustering progress leads to difficult model training especially in the initial stage, the possible feature space distortion, inaccurate and weak feature representation; (2) Existing deep clustering methods are difficult to intuitively define the similarity of time series and rely heavily on complex feature extraction networks and clustering algorithms. (3) Multidimensional time series have the characteristics of high dimensions, complex relationships between dimensions, and variable data forms, thus generating a huge feature space. It is difficult for existing methods to select discriminative features, resulting in generally low accuracy of methods. Accordingly, to address the above three problems, we proposed a novel general two-stage multi-dimensional spatial features based multi-view deep clustering method 1DCAE-TSSAMC (One-dimensional deep convolutional auto-encoder based two-stage stepwise amplification multi-clustering). We conducted verification and analysis based on real-world important multi-scenario, and compared with many other benchmarks ranging from the most classic approaches such as K-means and Hierarchical to the state-of-the-art approaches based on deep learning such as Deep Temporal Clustering (DTC) and Temporal Clustering Network (TCN). Experimental results show that the new method outperforms the other benchmarks, and provides more accurate, richer, and more reliable analysis results, more importantly, with significant improvement in accuracy and spatial linear separability.

Characterization of cancer survivors clustered by subjective and objective cognitive function scores.

Cancer-related cognitive impairment (CRCI) is a prevalent condition that significantly impacts the quality of life of individuals who receive cancer treatment. Clinical management of CRCI presents challenges due to the absenceof a standardized assessment. This study identified clinically relevant phenotypic clusters of CRCI based on subjective and objective cognitive function scores. In this cross-sectional study, participants were clustered using the VARCLUS™ based on subjective cognitive impairment assessed through the PROMIS® version 1.0 short-form subscales of cognitive abilities and cognitive concerns and the CANTAB Cambridge Cognition® scores, which included measures of visuospatial working memory capacity, visual episodic memory, new learning, working memory, executive function, and sustained attention. Each cluster's characteristics were described using demographics, physical and psychosocial factors (physical function, affect, optimism, and social support), and psychoneurological symptoms (anxiety, depression, fatigue, neuropathic pain, and sleep disturbance). We obtained five clusters from a total of 414 participants, where 99% were female, and 93% were self-reported white. Clusters 4 and 5 showed the highest PROMIS® cognitive abilities and the lowest measures of cognitive concern, while Clusters 1 and 2 showed the lowest cognitive abilities and the highest cognitive concerns. Clusters 4 and 5 had higher education, income, employment, and higher scores in physical function, positive affect, optimism, and social support. Additionally, individuals in these clusters were less prone to experience severe cancer-related psychoneurological symptoms. Our clustering approach, combining subjective and objective cognitive function information, shows promise in identifying phenotypes that hold clinical relevance for categorizing patient presentation of CRCI and facilitating individualized management strategies.

Domain Adaptation With Self-Supervised Learning and Feature Clustering for Intelligent Fault Diagnosis.

Domain adaptation indeed promotes the progress of intelligent fault diagnosis in industrial scenarios. The abundant labeled samples are not necessary. The identical distribution between the training and testing datasets is not any more the prerequisite for intelligent fault diagnosis working. However, two issues arise subsequently: Feature learning in domain adaptation framework tends to be biased to the source domain, and unreliable pseudolabeling seriously impacts on the conditional domain adaptation. In this article, a new domain adaptation approach with self-supervised learning and feature clustering (DASSL-FC) is proposed, trying to alleviate the issues by unbiased feature learning and pseudolabels updating strategy. Taking different transformation methods as pretext, the transformed data and its pretext train a neural network in an SSL way. As to pseudolabeling, clusters are taken as the auxiliary information to correct the network predicted labels in terms of the "strong cluster" rule. Then, the updated pseudolabels and their confidence are enforced to further estimate the conditional distribution discrepancy and its confidence weight. To verify the effectiveness of the proposed method, the experiments are implemented on intraplatform and interplatforms for simulating the practical scenarios.

Empowering Machine Learning Forecasting of Labquake Using Event‐Based Features and Clustering Characteristics

AbstractFollowing recent advances of machine learning (ML), we present a novel approach to extract spatiotemporal seismo‐mechanical features from Acoustic Emission (AE) catalogs to empower ML‐based forecasting. The AE data were recorded during laboratory stick‐slip experiments on granite samples cut by rough faults. Based on the features computed for a past time window, a random forest (RF) classifier is used to forecast the occurrence of a large magnitude event (MAE &gt; 3.5) in the next time window. Event‐based features allow us to associate informative time‐space characteristics to each feature and nearest‐neighbor clustering analysis enables us to separate background and clustered seismicity and train individual models. The results show that the separation of AEs enhances the forecasting accuracy from 73.2% for the entire catalog up to 82.1% and 89.0% if background and clustered events are used separately. The presented new approach may be upscaled for applications to forecast tectonic earthquakes.

The integrate profiling of single-cell and spatial transcriptome RNA-seq reveals tumor heterogeneity, therapeutic targets, and prognostic subtypes in ccRCC.

Clear-cell renal cell carcinoma (ccRCC) is the most common type of RCC; however, the intratumoral heterogeneity in ccRCC remains unclear. We first identified markers and biological features of each cell cluster using bioinformatics analysis based on single-cell and spatial transcriptome RNA-sequencing data. We found that gene copy number loss on chromosome 3p and amplification on chromosome 5q were common features in ccRCC cells. Meanwhile, NNMT and HILPDA, which are associated with the response to hypoxia and metabolism, are potential therapeutic targets for ccRCC. In addition, CD8+ exhausted T cells (LAG3+ HAVCR2+), CD8+ proliferated T cells (STMN+), and M2-like macrophages (CD68+ CD163+ APOC1+), which are closely associated with immunosuppression, played vital roles in ccRCC occurrence and development. These results were further verified by whole exome sequencing, cell line and xenograft experiments, and immunofluorescence staining. Finally, we divide patients with ccRCC into three subtypes using unsupervised cluster analysis. and generated a classifier to reproduce these subtypes using the eXtreme Gradient Boosting algorithm. Our classifier can help clinicians evaluate prognosis and design personalized treatment strategies for ccRCC. In summary, our work provides a new perspective for understanding tumor heterogeneity and will aid in the design of antitumor therapeutic strategies for ccRCC.

Ductility fracture mechanisms of Al-7Si-Mg casting alloys considering Si particles: A combined experimental and crystal plasticity study

Al-7Si-Mg casting alloys are widely utilized for their excellent formability, though they commonly exhibit poor ductility. This study comprehensively investigates the factors influencing the ductility of Al-7Si-Mg casting alloys, proposing a fracture mechanism and developing a model for ductile fracture of the alloy. Tensile tests were conducted on samples subjected to different aging times, with analysis of the alloy’s microstructure and tensile fracture surfaces. The findings reveal that fracture initiation occurs at clusters formed by the aggregation of Si particles, followed by the coalescence of cracks within these clusters, leading to macroscopic fracture. Different tensile strains were applied to specimens subjected to under-aging and over-aging treatments. The distribution of geometrically necessary dislocations (GNDs) in stretched samples indicates that the Si cluster area undergoes greater localized plastic deformation compared to the Al matrix, and this disparity increases with the applied strain. It has been observed that the reduced elongation observed in over-aged alloys is due to the coalescence of Si clusters under lower strain conditions. Representative volume elements with the characteristics of Si cluster is proposed, and the crystal plasticity finite element method is used for tensile simulations. The simulation results show that the proposed model is more consistent with experimental findings pertaining to elongation than with the results of a conventional void model. A longer aging time reduces ductility owing to a more intense strain localization under same tensile strain.

Feature extraction and clustering of electricity market data based on GCN and graph filtering

Feature extraction and clustering of electricity market data based on GCN and graph filtering