Cluster Centers Research Articles

Ferritin Level Prediction in Patients with Chronic Kidney Disease using Cluster Centers on Fuzzy Subtractive Clustering

Ferritin Level Prediction in Patients with Chronic Kidney Disease using Cluster Centers on Fuzzy Subtractive Clustering

The heart of galaxy clusters: Demographics and physical properties of cool-core and non-cool-core halos in the TNG-Cluster simulation

We analyzed the physical properties of the gaseous intracluster medium (ICM) at the center of massive galaxy clusters with TNG-Cluster, a new cosmological magnetohydrodynamical simulation. Our sample contains 352 simulated clusters spanning a halo mass range of 1014 < M500c/M⊙ < 2 × 1015 at z = 0. We focused on the proposed classification of clusters into cool-core (CC) and non-cool-core (NCC) populations, the z = 0 distribution of cluster central ICM properties, and the redshift evolution of the CC cluster population. We analyzed the resolved structure and radial profiles of entropy, temperature, electron number density, and pressure. To distinguish between CC and NCC clusters, we considered several criteria: central cooling time, central entropy, central density, X-ray concentration parameter, and density profile slope. According to TNG-Cluster and with no a priori cluster selection, the distributions of these properties are unimodal, whereby CCs and NCCs represent the two extremes. Across the entire TNG-Cluster sample at z = 0 and based on the central cooling time, the strong CC fraction is fSCC = 24%, compared to fWCC = 60% and fNCC = 16% for weak and NCCs, respectively. However, the fraction of CCs depends strongly on both halo mass and redshift, although the magnitude and even direction of the trends vary with definition. The abundant statistics of simulated high-mass clusters in TNG-Cluster enabled us to match observational samples and make a comparison with data. The CC fractions from z = 0 to z = 2 are in broad agreement with observations, as are the radial profiles of thermodynamical quantities, globally as well as when divided as CC versus NCC halos. TNG-Cluster can therefore be used as a laboratory to study the evolution and transformations of cluster cores due to mergers, AGN feedback, and other physical processes.

The Stability of Dense Cores near the Serpens South Protocluster

Most stars form in clusters and groups rather than in isolation. We present ≲5″ angular resolution (∼2000 au, or 0.01 pc) Very Large Array NH3 (1,1), (2,2), and (3,3) and 1.3 cm continuum emission observations of the dense gas within the Serpens South protocluster and extended filaments to the north and south. We identify 94 dense cores using a dendrogram analysis of the NH3 (1,1) integrated intensity. Gas temperatures T K and nonthermal line widths σ NT both increase toward the center of the young stellar cluster, in the dense gas generally and in the cores specifically. We find that most cores (54%) are supervirial, with gravitationally bound cores located primarily in the filaments. Cores in the protocluster have higher virial parameters by a factor of ∼1.7, driven primarily by the increased core σ NT values. These cores cannot collapse to form stars unless they accrete additional mass or their core internal motions are reduced. The southern filament shows a significant velocity gradient previously interpreted as mass flow toward the cluster. We find more complex kinematics in the northern filament. We find a strong correlation between σ NT and T K , and argue that the enhanced temperatures and nonthermal motions are due to mechanical heating and interaction between the protocluster-driven outflows and the dense gas. Filament-led accretion may also contribute to the increased σ NT values. Assuming a constant fraction of core mass ends up in the young stars, future star formation in the Serpens South protocluster will shift to higher masses by a factor of ∼2.

Clustering statistical method of high dimensional sparse data based on fuzzy data

Abstract Developing effective clustering and statistical methods for high-dimensional sparse data presents unique challenges compared to traditional low-dimensional data. To address this, a novel approach is proposed, leveraging fuzzy data principles to enhance the clustering and statistical performance of high-dimensional sparse datasets. The method builds upon the fuzzy C-means clustering algorithm, introducing key modifications for better suitability to high-dimensional sparse data. One crucial enhancement involves tackling the local optimization problem by optimizing the initial clustering center, significantly reducing clustering statistical time. Replacing the original Euclidean distance with cosine distance improves the clustering and statistical performance of high-dimensional sparse data. Experimental results have shown that this method has superior clustering statistical performance when the data dimensions are different. When the data dimension is low, and the blocking ratio is 10%, the clustering statistical effect is optimal. When the data dimension is high, and the blocking ratio is 40%, the clustering statistical effect is optimal. This method has higher hit rates and clustering statistical efficiency at different sparsity levels.

Screening of Drought-Resistant Potato Germplasm Resources Based on Intelligent Algorithm

As a staple food resource, potato is of great significance for improving grain reserves and ensuring national food security. In order to improve potato yield and promote the process of potato becoming a staple food, a K-means algorithm optimized by particle swarm algorithm was proposed to realize the screening of dry potato germplasm resources. First, the research continues the research on particle swarm optimization, and innovatively applies K-means algorithm to optimization. The research utilizes the advantages of particle swarm optimization, such as fast convergence speed, strong search ability, and simple operation, to enable particle swarm optimization to take on the role of optimizing the initial clustering center, thereby improving the accuracy and efficiency of clustering analysis. On this basis, a PSO-K-means drought resistant potato germplasm resource screening model was constructed. This model consists of a data collection and preprocessing module, an impact indicator determination module, and a comprehensive evaluation module. Finally, the application effect of the model was verified. The results show that the AUC value of the model is up to 0.840, and the screening accuracy is as high as 94.5%, which is 13.5% higher than that of the K-means model. The research method has been validated to improve the limitations of K-means mode, such as high screening error, weak stability, and falling into local optimal solutions. It optimizes the screening effect of drought resistant potato germplasm resources, which is conducive to exploring the potential of potato resources. In addition, research has also provided broader ideas for the optimization and application of particle swarm optimization algorithms.

Solution of the structure of the high-coverage CO layer on the Ru(0001) surface—A combined study by density functional theory and scanning tunneling microscopy

Structures formed by dense CO adsorption layers can provide information about the balance between molecule–surface and molecule–molecule interactions. However, in many cases, the structure models are not clear. Using density functional theory (DFT) and scanning tunneling microscopy (STM), we have investigated the high-coverage CO layer on the Ru(0001) surface. Previous investigations by low-energy electron diffraction (LEED) and vibrational spectroscopy led to conflicting results about the structure. In the present study, 88 models with coverages between 0.58 and 0.77 monolayers have been analyzed by DFT. The most stable structures consist of small, compact CO clusters with an internal pseudo 1×1 structure. The CO molecules in the cluster centers occupy on-top sites in an upright position, whereas the molecules farther outside are slightly shifted from these sites and tilted outward. STM data of the CO-saturated surface at low temperatures, corresponding to a coverage of 0.66 monolayers, show a quasi-hexagonal pattern of features with an internal hexagonal fine structure. Simulated images based on the cluster model agree with the experimental data. It is concluded that the high-coverage CO layer consists of the close-packed clusters predicted by DFT as the most stable structure elements. In the experiment, the sizes and shapes of the clusters vary. However, the arrangement is not random but follows defined tiling rules. The structure remains ordered, almost up to room temperature. The LEED data are re-interpreted on the basis of the Fourier transforms of the STM data, solving the long-standing conflict about the structure.

PointStaClu: A Deep Point Cloud Clustering Method Based on Stable Cluster Discrimination

Potential inconsistencies between the goals of unsupervised representation learning and clustering within multi-stage deep clustering can diminish the effectiveness of these techniques. However, because the goal of unsupervised representation learning is inherently flexible and can be tailored to clustering, we introduce PointStaClu, a novel single-stage point cloud clustering method. This method employs stable cluster discrimination (StaClu) to tackle the inherent instability present in single-stage deep clustering training. It achieves this by constraining the gradient descent updates for negative instances within the cross-entropy loss function, and by updating the cluster centers using the same loss function. Furthermore, we integrate entropy constraints to regulate the distribution entropy of the dataset, thereby enhancing the cluster allocation. Our framework simplifies the process, employing a single loss function and an encoder for deep point cloud clustering. Extensive experiments on the ModelNet40 and ShapeNet dataset demonstrate that PointStaClu significantly narrows the performance gap between unsupervised point cloud clustering and supervised point cloud classification, presenting a novel approach to point cloud classification tasks.

중원지역 원삼국~삼국시대 중심취락의 변동과 지역사회

This paper is written with the aim of classifying the settlements in the Chungju area of the Central Plains region, utilizing the recently introduced Social Network Analysis (SNA) in Korean archaeology, and examining the transitions of settlements and regional characteristics during the transition from the Proto-Three Kingdoms to the Three Kingdoms Era. Settlements in the Chungju area have the disadvantage of being numerically smaller than those in Cheongju. Therefore, the periodization is based on the expansion periods of Baekje, Goguryeo, and Silla: Phase I (before the mid-4th century), PhaseⅡ(from the mid-4th century to the late 6th century), and PhaseⅢ(after the mid-6th century). During Phase I, Geumneung-dong functioned as the central settlement with Tangeumdae-Fortress and Chilgeum-dong playing significant roles in the Mahan Central Settlement. During PhaseⅡ, with the expansion of Baekje, Tappyeong-ri emerged as a central settlement, utilizing the alluvial lands and hillsides for residential and burial purposes. Tangeumdae Fortress and Chilgeum-dong continued their roles as iron production sites from Phase I. Meanwhile, Baekje left its identity in the Hwangsae-maeri tomb cluster. In PhaseⅡ, iron production by Baekje expanded east to west from Tappyeong-ri. Despite the Goguryeo expansion, Tappyeong-ri continued to serve as a central hub. Additionally, evidence of Goguryeo-related tombs and residences in Dujeong-ri and Danwol-dong suggests a prolonged Goguryeo occupation in the Chungju area. However, in PhaseⅢ, Chungju witnessed the expansion of Silla. Silla utilized Tappyeong-ri, previously occupied by Baekje and Goguryeo, as its central settlement and established a central tomb cluster in Nuam-ri and Haguam-ri. Recently reported remains at 445-1 Tappyeong-ri indicate a broader extent of settlement in Tappyeong-ri, suggesting its continued significance as a central settlement even after the Silla expansion. During this period, surrounding settlements around Tappyeong-ri, Nuam-ri, and Haguam-ri expanded even further compared to Phases I andⅡ. However, in the late 7th century of PhaseⅢ, Danwol-dong emerged prominently, indicating a shift of the central hub towards the east. In essence, the central settlements in the Chungju area evolved from Geumneung-dong, Tangeumdae, and Chilgeum-dong (East) to Tappyeong-ri, Nuam-ri, and Haguam-ri (West), and finally to Danwol-dong (East). This transformation correlates with the expansion processes of the Three Kingdoms era following the Proto-Three Kingdoms period. However, there were differences in the formation and positioning of the central settlements among the Three Kingdoms. Specifically, Baekje chose locations reflecting intentions to monitor or exploit nearby Mahan forces, while Goguryeo and Silla directly occupied Tappyeong-ri. This is interpreted as a measure to counteract the weakening and containment of previous powers.

Clustering-based interactive image segmentation

Objectives. The task of color image segmentation without the use of preliminary training is considered. It arises, for example, when it is necessary to perform image segmentation with semantic and color properties unknown in advance immediately after their acquisition, or when the set of images intended for segmentation is too small, as well as when performing preliminary "exploratory" analysis of images. In such cases, powerful neural network and other segmentation tools that require deep learning can not be used.Methods. An algorithm for interactive image segmentation is proposed, based on the analysis of the colors of areas selected interactively. First, in interactive mode, the image areas belonging to the objects are selected very approximately, and then regions belonging to the background are chosen. In the next step, the set of colors of the selected object areas and the set of colors of the selected background areas are clustered separately by one of the clustering algorithms, for example, k-means, fuzzy c-means, or the multi-level clustering algorithm proposed by the author. After this, non-informative elements are removed from the set of cluster centers describing the objects and the set of clusters presenting the background. The modified sets of object and background cluster centers are used for image segmentation.Results . The constructed algorithm allows selection of the required objects in color images if the colors of the objects and the background are different. Interactive selection of object areas and background areas does not require accuracy or much effort and usually takes several tens of seconds. For selection, rectangular areas that lie entirely inside the object images, and rectangular areas that belong completely to the background can be used. Below an example of interactive regions selection and color image segmentation is shown.Conclusion. The experiments performed showed the effectiveness of the proposed approach to segmenting color images. It can be used in cases where the semantic and color properties of images are not known in advance, and in cases where the use of more powerful deep learning methods, including neural networks, is too expensive or impossible.

Research on acoustic localization method for multiple leakage sources of water-cooling wall in furnace

Abstract The water-cooling wall of the furnace often leaks under harsh conditions such as high temperature and pressure, which affects the safety and economic operation of the boiler. To achieve the localization of multiple leakage sources in water-cooling wall, an acoustic positioning method based on improved differential evolution and density-based spatial clustering of application with noise (IDE-DBSCAN) is proposed. The nonlinear equation system is constructed using time delay of arrival (TDOA) obtained by the generalized second cross-correlation algorithm, and the IDE algorithm is used to solve nonlinear equation system and obtain multiple spatial positions corresponding to multiple TDOA sequences. To eliminate false leakage sources, the paper presents IDE-DBSCAN to cluster the multiple estimated positions. The number of clusters obtained is the number of leakage sources, and the cluster centers are the location coordinates of the leakage sources. The numerical simulation results show that compared to traditional acoustic positioning methods, IDE-DBSCAN has higher positioning accuracy and better anti-noise ability. In addition, experiments are implemented on multiple leakage sound sources within a measurement area of 1m×1m×1.06m, and the positioning errors remain within 5cm. The proposed method provides a theoretical basis and experimental verification for further research on multiple leakages localization of water-cooling wall in furnace.

Comparative Study on the Agglomeration Degree and Influencing Factors of the Urban Creative Class in the Central Area of the Yangtze River Delta

The creative class has become an important force in promoting sustainable urban development. Existing research has explored many factors for the agglomeration of the creative class, but there is still a lack of comparative studies from a heterogeneity perspective and on the design of cross-cultural adaptability factors, especially comparative studies across different regions in Chinese cities. This paper focuses on 27 central district cities in the Yangtze River Delta region of China, based on panel data from 2012 to 2019, and uses the location entropy method to measure the agglomeration degree of the creative class in different cities. Considering the cultural background of China, a model of influencing factors including five dimensions: economic foundation, living environment, cultural and educational environment, innovation environment, and ecological environment is constructed. The study finds: (1) The agglomeration degree of the creative class in the central urban cluster of the Yangtze River Delta is significantly higher than the peripheral level, forming an agglomeration distribution pattern led by Shanghai, with Nanjing, Hefei, Hangzhou, Suzhou, Ningbo, and other important node cities. The agglomeration degree of the creative class generally shows an upward trend. (2) The economic foundation, living environment, cultural and educational environment, innovation environment, and ecological environment all have a significant positive impact on the agglomeration of the creative class, among which the impact of the living environment, cultural and educational environment, and innovation environment is larger, while the economic foundation and ecological environment are relatively smaller. (3) There are differences in the influencing factors of the creative class agglomeration at the provincial levels of Jiangsu, Zhejiang, and Anhui in the Yangtze River Delta. Compared with the more core provinces of Jiangsu and Zhejiang, the cultural and educational environment and ecological environment have a more significant impact, while the relatively peripheral Anhui province is more dependent on the city’s openness and innovation environment factors. This study reveals the spatial distribution rules and influencing factors of the creative class in the central cities of the Yangtze River Delta, providing insights for coordinated and sustainable regional development.

Effect of droplet spacing on micro-explosion and combustion characteristics of multi-component fuel droplet cluster

The micro-explosion and combustion characteristics of n-butanol/n-pentanol/biodiesel blended fuel droplet cluster are investigated using high-speed backlit imaging technique. The droplet cluster includes 9 droplets with the same diameter, which are arranged in a 3 × 3 matrix equidistantly using the fiber support method. The results indicate that when the normalized droplet spacing (s/d0) is less than 3, the coalescence phenomenon caused by micro-explosions will occur between adjacent two or three droplets in the droplet cluster. For small s/d0 (2 ∼ 4.5), the droplet cluster sustains a fully enveloped flame. Within the range of s/d0 of 5 ∼ 7, two combustion modes emerge: transitioning from a fully enveloped flame to an independent flame, and merging independent flames into enveloped flames, eventually reverting to independent flames. At a s/d0 of 8, each droplet within the cluster maintains an independent flame. In addition, the combustion rate of droplets 1–3 and 7–9 reaches its maximum value at 4 ∼ 4.5 s/d0, about 30 % higher than single droplet. Most importantly, the droplet micro-explosion intensity at the center of the droplet cluster is higher than that of other droplets (similar to a single droplet), which is caused by the “group micro-explosion” effect.

An Improved YOLOv5x-Based Algorithm for IC Pin Welding Defects Detection

This study suggests an integrated circuit (IC) pin welding defect detection algorithm based on improved YOLOv5x to address the issues of low detection accuracy caused by small target size and dense pin arrangement in IC pin welding defects identification. The ability of the network to extract features is improved by effective fusing of features with various receptive fields through the inclusion of the D-SPP module to merge different channel information. The introduction of the mask self-attention mechanism module increases the network’s capacity to recognize global feature correlations and raises the algorithm’s detection precision. In order to speed up the model’s convergence and tackle the issue of sample imbalance in BBox regression, the Focal-EIoU loss function is applied. The detection accuracy and speed are increased by using the k-means++ clustering algorithm to create nine clustering centers to figure out the size of the prior box. According to the results of the experiment, the new method achieves average precisions for short-circuit, missing pin, pin-cocked, and little tin faults in IC pin welding of 96.7%, 94.5%, 95.6%, and 93.3%, respectively. The mean average precision increases to 95.0% at a threshold of 0.5, which is 13.3% and 8.9% greater than YOLOv3 and YOLOv5x, respectively. A reference value for IC pin welding defect identification is provided by the improved algorithm, which has a detection time of 0.142 seconds per image. This meets the speed requirements of IC quality inspection.

Research on the classification of ancient silicate glass artifacts based on machine learning

AbstractClassifying cultural relics has always been a major challenge for archaeologists. Using glass artifacts as the research object, a classification model for glass artifacts was constructed using decision trees, support vector machines, and logistic regression methods based on their patterns, colors, surface weathering conditions, types, and composition ratios. Three models were used to identify the types of unknown glass artifacts. A subclassification model for high‐potassium glass and lead barium glass was established using the K‐means clustering method. The elbow method and average contour method were used to determine the optimal number of clusters, and the decision tree model was named based on the characteristics of the cluster center components. The research results indicate that the three models yield consistent identification results for unknown types of glass relics, and the classification results are good. Lead barium glass and high‐potassium glass can be divided into three and six subclasses, respectively, and the naming of the subclass decision tree is reasonable. The identification method for ancient glass relics in this article is highly practical and can provide a reference for the classification and identification of other component data.

A Novel Approach for Mobility-Aware Energy-Efficient Clustering-Based Routing Using EANN With GA Algorithm on WSNs

Aim: This research aims to explore and evaluate various strategies for improving energy efficiency within wireless sensor networks (WSNs). Specifically, the study focuses on the critical challenge of extending network lifespan through energy conservation by establishing balanced clusters within the WSN architecture. Background: In wireless sensor networks (WSNs), ensuring prolonged network operation while conserving energy resources is a significant concern. One promising approach to address this challenge is the implementation of equalized clusters, which requires an effective selection of cluster heads (CHs). However, this task presents considerable complexity and demands innovative solutions to overcome. Objective: The primary objective of this study is to develop and assess a novel methodology for selecting precise cluster heads (CHs) within WSNs. This methodology is based on the utilization of Bluetooth low energy (BLE) sensors deployed in a randomly distributed manner across the study area. By employing an enhanced artificial neural network and greedy approach (EANN-GA), the proposed technique seeks to identify CHs with optimal proximity to the cluster center and substantial remaining energy reserves. Methods: The proposed methodology involves the deployment of BLE sensors distributed randomly throughout the study region, which are then organized into clusters. Using the enhanced artificial neural network and greedy approach (EANN-GA), the sensor node nearest to the cluster center with the highest remaining energy is selected as the cluster head (CH). Additionally, a mobile sink (MS) is introduced to harness the power of CHs, and the number of paths utilized by the MS is estimated through a genetic approach. Based on this path information, the MS enters each cluster to initiate the data-gathering process. Results: Performance analysis of the presented methodology demonstrates significant improvements in energy efficiency and the extension of network lifetime. By employing the proposed EANN-GA technique for CH selection and optimizing MS path utilization, the study showcases enhanced operational effectiveness within WSNs. Conclusion: The findings of this research underscore the effectiveness of the proposed methodology in enhancing energy efficiency and prolonging the lifespan of wireless sensor networks. Through the innovative integration of BLE sensors, EANN-GA CH selection, and genetic-based MS path estimation, the study contributes valuable insights toward addressing the critical challenges of energy conservation in WSNs.

Visual Field Prognosis From Macula and Circumpapillary Spectral Domain Optical Coherence Tomography.

To explore the structural-functional loss relationship from optic-nerve-head- and macula-centred spectral-domain (SD) Optical Coherence Tomography (OCT) images in the full spectrum of glaucoma patients using deep-learning methods. A cohort comprising 5238 unique eyes classified as suspects or diagnosed with glaucoma was considered. All patients underwent ophthalmologic examination consisting of standard automated perimetry (SAP), macular OCT, and peri-papillary OCT on the same day. Deep learning models were trained to estimate G-pattern visual field (VF) mean deviation (MD) and cluster MD using retinal thickness maps from seven layers: retinal nerve fiber layer (RNFL), ganglion cell layer and inner plexiform layer (GCL + IPL), inner nuclear layer and outer plexiform layer (INL + OPL), outer nuclear layer (ONL), photoreceptors and retinal pigmented epithelium (PR + RPE), choriocapillaris and choroidal stroma (CC + CS), total retinal thickness (RT). The best performance on MD prediction is achieved by RNFL, GCL + IPL and RT layers, with R2 scores of 0.37, 0.33, and 0.31, respectively. Combining macular and peri-papillary scans outperforms single modality prediction, achieving an R2 value of 0.48. Cluster MD predictions show promising results, notably in central clusters, reaching an R2 of 0.56. The combination of multiple modalities, such as optic-nerve-head circular B-scans and retinal thickness maps from macular SD-OCT images, improves the performance of MD and cluster MD prediction. Our proposed model demonstrates the highest level of accuracy in predicting MD in the early-to-mid stages of glaucoma. Objective measures recorded with SD-OCT can optimize the number of visual field tests and improve individualized glaucoma care by adjusting VF testing frequency based on deep-learning estimates of functional damage.

Exploring road safety using alignment perspective features in real driving images: A case study on mountain freeways.

While driving, drivers frequently adapt their driving behaviors according to their perception of the road's alignment features. However, traditional two-dimensional alignment methods lack the ability to capture these features from the driver's perspective. This study introduces a novel method for road alignment recognition, employing image recognition technology to extract alignment perspective features, namely alignment perspective skewness (APS) and alignment perspective kurtosis (APK), from in-real driving images. Subsequently, the K-means clustering algorithm is utilized for road segment classification based on APS and APK indicators. Various sliding step length for clustering are employed, with step length ranging from 100m to 400m. Furthermore, the accident rates for different segment clusters are analyzed to explore the relationship between alignment perspective features and traffic safety. A 150 km mountain road section of the Erlianhaote-Guangzhou freewway from Huaiji to Sihui is selected as a case study. The results demonstrate that using alignment perspective features as classification criteria produces favorable clustering outcomes, with superior clustering performance achieved using shorter segment lengths and fewer cluster centers. The road segment classification based on alignment perspective features reveals notable differences in accident rates across categories; while traditional two-dimensional parameters-based classification methods fail to capture these differences. The most significant differences in accident rates across categories are observed with segment length of 100m, with the significance gradually diminishing as segment length increases and disappearing entirely when the length exceeds 300m. These findings validate the reliability of using alignment perspective features (APS and APK) for road alignment classification and road safety analysis, providing valuable insights for road safety management.

Multiple Constraint Hybrid Travel Route Recommendation Model Based on Collaborative Filtering Recommendation Algorithm

The study presents a hybrid model for recommending travel routes that takes into account multiple constraints. This model is based on a collaborative filtering recommendation algorithm and addresses the issue of disorganized travel route recommendations. To improve upon the k-mean clustering algorithm, the proposed model introduces the Dynamic Gaussian Kernel Density K-means algorithm. After the data was processed, the initial clustering center was determined and k-means clustering was performed. Subsequently, the travel route recommendation model was created by integrating various constraints. The study’s proposed algorithm was compared with alternative algorithms, and the experimental results demonstrated superior performance across a range of datasets, with the minimum sum of squared errors and a running time of approximately 1.4 seconds - a noteworthy improvement. Comparative experiments were conducted on various forgetting coefficients in the model, and the forgetting coefficient with the lowest sum-of-squares of errors was selected to replace the existing one. Upon comparing the proposed research model with other models, it was found that the former had greater accuracy and recall, amounting to 98.1% and 96.8% respectively. This suggests that the proposed research model serves as a more efficient solution for travel route recommendation.

Dual space-based fuzzy graphs and orthogonal basis clustering for unsupervised feature selection

Unsupervised feature selection (UFS) takes an important position because gaining the class labels is laborious or even impossible. In the domain of UFS, clustering is a major means to exploit label information. The existing methods either could not model a clear clustering structure or could not utilize the local structure of data. Consequently, a new clustering method in combination with graph learning is proposed in this work. Specifically, for clustering, orthogonal basis clustering is introduced, where orthogonal constraints are imposed on the cluster center matrix and the clustering matrix. The clustering indicator matrix is also imposed by a non-negative constraint. A clearer clustering structure and more independent clustering centers are obtained through these constraints. For local preservation, given that traditional graphs for keeping the local manifold are faced with the problem of imbalanced neighbors, the fuzzy graph is introduced to acquire a robust structure, which is applied to both data space and feature space. The topological structure in these spaces can be well maintained. For the choice of salient features, ℓ2,0-norm regularization is imposed on the projection matrix. The object function is solved alternately. Then, a feature selection algorithm is designed. Experiments are designed and performed on nine real-world data sets. The results attest to the effectiveness of the proposed algorithm compared with other relevant algorithms.

Research on color and texture characteristics and visual perception of custom wardrobe finishes

In recent years, the custom wardrobe market has been steadily developing. While meeting the functional needs of users, it is gradually shifting towards aesthetic preferences. Rapidly grasping users’ preferences for the appearance of custom wardrobes is a key focus of current research. This study collected a large number of decorative surface images of custom wardrobes and objectively analyzed the design features based on color moments and Tamura texture feature data in computer image analysis methods. K-means cluster analysis was performed on the feature data. Collected images of the points closest to the cluster centers were further screened to select representative finish images, and finally a questionnaire survey was conducted at Nanjing Forestry University, with the help of semantic differential method and factor analysis. The characteristics of the samples were comprehensively summarized to infer design elements. The study found that warm-toned, medium-low saturation, and medium brightness surfaces were preferred by the panel. Different colors, contrasts, saturations, brightness, element features, and arrangements have significantly different effects on visual perception. These conclusions can provide a reference for subsequent custom wardrobe design.