RESUMO
Depending on the node ordering, an adjacency matrix can highlight distinct characteristics of a graph. Deriving a "proper" node ordering is thus a critical step in visualizing a graph as an adjacency matrix. Users often try multiple matrix reorderings using different methods until they find one that meets the analysis goal. However, this trial-and-error approach is laborious and disorganized, which is especially challenging for novices. This paper presents a technique that enables users to effortlessly find a matrix reordering they want. Specifically, we design a generative model that learns a latent space of diverse matrix reorderings of the given graph. We also construct an intuitive user interface from the learned latent space by creating a map of various matrix reorderings. We demonstrate our approach through quantitative and qualitative evaluations of the generated reorderings and learned latent spaces. The results show that our model is capable of learning a latent space of diverse matrix reorderings. Most existing research in this area generally focused on developing algorithms that can compute "better" matrix reorderings for particular circumstances. This paper introduces a fundamentally new approach to matrix visualization of a graph, where a machine learning model learns to generate diverse matrix reorderings of a graph.
RESUMO
The rapid development of Convolutional Neural Networks (CNNs) in recent years has triggered significant breakthroughs in many machine learning (ML) applications. The ability to understand and compare various CNN models available is thus essential. The conventional approach with visualizing each model's quantitative features, such as classification accuracy and computational complexity, is not sufficient for a deeper understanding and comparison of the behaviors of different models. Moreover, most of the existing tools for assessing CNN behaviors only support comparison between two models and lack the flexibility of customizing the analysis tasks according to user needs. This paper presents a visual analytics system, VAC-CNN (Visual Analytics for Comparing CNNs), that supports the in-depth inspection of a single CNN model as well as comparative studies of two or more models. The ability to compare a larger number of (e.g., tens of) models especially distinguishes our system from previous ones. With a carefully designed model visualization and explaining support, VAC-CNN facilitates a highly interactive workflow that promptly presents both quantitative and qualitative information at each analysis stage. We demonstrate VAC-CNN's effectiveness for assisting novice ML practitioners in evaluating and comparing multiple CNN models through two use cases and one preliminary evaluation study using the image classification tasks on the ImageNet dataset.
Assuntos
Gráficos por Computador , Redes Neurais de Computação , Aprendizado de MáquinaRESUMO
Previous efforts to discover new surrogate markers for the central nervous system (CNS) inflammatory demyelinating disorders have shown inconsistent results; moreover, supporting evidence is scarce. The present study investigated the IgG autoantibody responses to various viral and autoantibodies-related peptides proposed to be related to CNS inflammatory demyelinating disorders using the peptide microarray method. We customized a peptide microarray containing more than 2440 immobilized peptides representing human and viral autoantigens. Using this, we tested the sera of patients with neuromyelitis optica spectrum disorders (NMOSD seropositive, n = 6; NMOSD seronegative, n = 5), multiple sclerosis (MS, n = 5), and myelin-oligodendrocyte glycoprotein antibody-associated disease (MOGAD, n = 6), as well as healthy controls (HC, n = 5) and compared various peptide immunoglobulin G (IgG) responses between the groups. Among the statistically significant peptides based on the pairwise comparisons of IgG responses in each disease group to HC, cytomegalovirus (CMV)-related peptides were most clearly distinguishable among the study groups. In particular, the most significant differences in IgG response were observed for HC vs. MS and HC vs. seronegative NMOSD (p = 0.064). Relatively higher IgG responses to CMV-related peptides were observed in patients with MS and NMOSD based on analysis of the customized peptide microarray.
RESUMO
Urban green space is thought to contribute to citizen happiness by promoting physical and mental health. Nevertheless, how urban green space and happiness are related across many countries with different socioeconomic conditions has not been explored. By measuring the urban green space score (UGS) from high-resolution satellite imagery of 90 global cities covering 179,168 km2 and 230 million people in 60 developed countries, we find that the amount of urban green space and GDP are correlated with a nation's happiness level. More specifically, urban green space and GDP are each individually associated with happiness. Yet, only urban green space is related to happiness in the 30 wealthiest countries, whereas GDP alone can explain happiness in the subsequent 30 countries in terms of wealth. We further show that the relationship between urban green space and happiness is mediated by social support and that GDP moderates this relationship. These findings corroborate the importance of maintaining urban green space as a place for social cohesion to support people's happiness. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1140/epjds/s13688-021-00278-7.
RESUMO
BACKGROUND: Neuromyelitis optica spectrum disorder (NMOSD) targets astrocytes and elevates the levels of astrocyte-injury markers during attacks. FAM19A5, involved in reactive gliosis, is secreted by reactive astrocytes following central nervous system (CNS) damage. OBJECTIVE: To investigate the significance of serum FAM19A5 in patients with NMOSD. METHODS: We collected clinical data and sera of 199 patients from 11 hospitals over 21 months. FAM19A5 levels were compared among three groups: NMOSD with positive anti-aquaporin-4 antibody (NMOSD-AQP4), other CNS demyelinating disease, and healthy controls. RESULTS: The median serum FAM19A5 level was higher in the NMOSD-AQP4 (4.90 ng/mL (3.95, 5.79)) than in the other CNS demyelinating (2.35 ng/mL (1.83, 4.07), p < 0.001) or healthy control (1.02 ng/mL (0.92, 1.14), p < 0.001) groups. There were significant differences in the median serum FAM19A5 levels between the attack and remission periods (5.89 ng/mL (5.18, 6.98); 4.40 ng/mL (2.72, 5.13), p < 0.001) in the NMOSD-AQP4 group. Sampling during an attack (p < 0.001) and number of past attacks (p = 0.010) were independently associated with increased serum FAM19A5. CONCLUSION: Serum FAM19A5 was higher in patients with NMOSD-AQP4 and correlated with clinical characteristics. Thus, serum FAM19A5 may be a novel clinical biomarker for NMOSD-AQP4.
Assuntos
Neuromielite Óptica , Aquaporina 4 , Autoanticorpos , Biomarcadores , Humanos , Glicoproteína Mielina-Oligodendrócito , Neuromielite Óptica/diagnósticoRESUMO
Dimensionality reduction (DR) is frequently used for analyzing and visualizing high-dimensional data as it provides a good first glance of the data. However, to interpret the DR result for gaining useful insights from the data, it would take additional analysis effort such as identifying clusters and understanding their characteristics. While there are many automatic methods (e.g., density-based clustering methods) to identify clusters, effective methods for understanding a cluster's characteristics are still lacking. A cluster can be mostly characterized by its distribution of feature values. Reviewing the original feature values is not a straightforward task when the number of features is large. To address this challenge, we present a visual analytics method that effectively highlights the essential features of a cluster in a DR result. To extract the essential features, we introduce an enhanced usage of contrastive principal component analysis (cPCA). Our method, called ccPCA (contrasting clusters in PCA), can calculate each feature's relative contribution to the contrast between one cluster and other clusters. With ccPCA, we have created an interactive system including a scalable visualization of clusters' feature contributions. We demonstrate the effectiveness of our method and system with case studies using several publicly available datasets.
RESUMO
Different layouts can characterize different aspects of the same graph. Finding a "good" layout of a graph is thus an important task for graph visualization. In practice, users often visualize a graph in multiple layouts by using different methods and varying parameter settings until they find a layout that best suits the purpose of the visualization. However, this trial-and-error process is often haphazard and time-consuming. To provide users with an intuitive way to navigate the layout design space, we present a technique to systematically visualize a graph in diverse layouts using deep generative models. We design an encoder-decoder architecture to learn a model from a collection of example layouts, where the encoder represents training examples in a latent space and the decoder produces layouts from the latent space. In particular, we train the model to construct a two-dimensional latent space for users to easily explore and generate various layouts. We demonstrate our approach through quantitative and qualitative evaluations of the generated layouts. The results of our evaluations show that our model is capable of learning and generalizing abstract concepts of graph layouts, not just memorizing the training examples. In summary, this paper presents a fundamentally new approach to graph visualization where a machine learning model learns to visualize a graph from examples without manually-defined heuristics.
RESUMO
Using different methods for laying out a graph can lead to very different visual appearances, with which the viewer perceives different information. Selecting a "good" layout method is thus important for visualizing a graph. The selection can be highly subjective and dependent on the given task. A common approach to selecting a good layout is to use aesthetic criteria and visual inspection. However, fully calculating various layouts and their associated aesthetic metrics is computationally expensive. In this paper, we present a machine learning approach to large graph visualization based on computing the topological similarity of graphs using graph kernels. For a given graph, our approach can show what the graph would look like in different layouts and estimate their corresponding aesthetic metrics. An important contribution of our work is the development of a new framework to design graph kernels. Our experimental study shows that our estimation calculation is considerably faster than computing the actual layouts and their aesthetic metrics. Also, our graph kernels outperform the state-of-the-art ones in both time and accuracy. In addition, we conducted a user study to demonstrate that the topological similarity computed with our graph kernel matches perceptual similarity assessed by human users.
Assuntos
Algoritmos , Gráficos por Computador , Aprendizado de Máquina , Adolescente , Adulto , Estética , Feminino , Humanos , Masculino , Análise e Desempenho de Tarefas , Adulto JovemRESUMO
Information visualization has traditionally limited itself to 2D representations, primarily due to the prevalence of 2D displays and report formats. However, there has been a recent surge in popularity of consumer grade 3D displays and immersive head-mounted displays (HMDs). The ubiquity of such displays enables the possibility of immersive, stereoscopic visualization environments. While techniques that utilize such immersive environments have been explored extensively for spatial and scientific visualizations, contrastingly very little has been explored for information visualization. In this paper, we present our considerations of layout, rendering, and interaction methods for visualizing graphs in an immersive environment. We conducted a user study to evaluate our techniques compared to traditional 2D graph visualization. The results show that participants answered significantly faster with a fewer number of interactions using our techniques, especially for more difficult tasks. While the overall correctness rates are not significantly different, we found that participants gave significantly more correct answers using our techniques for larger graphs.
RESUMO
[Purpose] This study examined the effects of ramp slope (1:12, 1:10, 1:8, and 1:6) on physiological characteristics and performance times of wheelchair users and the performance times of caregivers to determine which slope would be the best for wheelchairs, in order to propose a ramp slope that incorporates a universal design. [Subjects and Methods] Twenty-four healthy subjects were enrolled in this study. Fifteen of these subjects also volunteered to participate as caregivers. A wooden ramp with an adjustable slope was constructed. As manual wheelchair users, the participants performed propulsion of a wheelchair up the ramp at a self-selected pace. Four ramp slopes (1:12, 1:10, 1:8, and 1:6) were used, and the participants sequentially ascended them in order from the gentlest to the steepest slope. The caregivers also pushed a wheelchair up the ramp at a self-selected pace. The blood pressure and pulse of participants after the ascent, as well as the performance times of the caregivers and manual wheelchair users, were measured on each of the different ramp slopes. The measured data, pulse, blood pressure, and performance time, were analyzed using repeated ANOVA. [Results] Systolic blood pressure was significantly higher after ascending the 1:6 slope than after ascending the 1:12 and 1:8 slopes. Diastolic blood pressure was significantly higher after ascending the 1:6 slope than after ascending the 1:12 and 1:8 slopes. The participants' pulses tended to increase significantly with an increase in slope. An assessment of the propulsion performance times revealed significant differences among the slopes. [Conclusion] Considering the results of the wheelchair users and caregivers, the 1:12 and 1:10 slopes are suitable ramp slopes for wheelchairs.
RESUMO
Argonaute 2 (Ago2) is a pivotal regulator of cell fate in adult stem cells. Its expression is significantly downregulated in late passages of cells, concomitant with a prominent increase in Ago2 cytosolic localization in single cells. Nuclear localization of Ago2 is crucial for the survival, proliferation, and differentiation of hATSCs (human adipose tissue-derived stem cells), mediated by the specific binding of the regulatory regions of functional genes, which positively or negatively altered gene expression. Ago2 targets genes that control stemness, reactive oxygen species scavenging, and microRNA expression, all of which are crucial for hATSC survival and self-renewal. Ago2 promotes cell proliferation and self-renewal by activating the expression of octamer-binding transcription factor 4 (Oct4). We confirmed the direct regulation of Oct4 activity by Ago2, as indicated by the results of the ChIP analysis. Methyl-CpG-binding protein 6 (MBD6) was detected as an Oct4 regulatory gene. As predicted, knockdown of MBD6 expression attenuated cell proliferation and eventually induced cell death. We hypothesized that MBD6 functions downstream of Oct4 in the regulation of stemness-related genes, cell proliferation, self-renewal activity, and survival. MBD6 also promoted cell transdifferentiation into neural and endodermal ß-cells while significantly attenuating differentiation into the mesodermal lineage. We demonstrate that MBD6 is regulated by Ago2 via an interaction with Oct4, which alters self-renewal and gene expression in hATSCs. MBD6 was promoted cell proliferation through a novel set of signal mediators that may influence differentiation by repressing MBD2 and MBD3, which are possibly recruited by germ cell nuclear factor (GCNF).
Assuntos
Tecido Adiposo/citologia , Células-Tronco Adultas/citologia , Proteínas de Ligação a DNA/metabolismo , Fator 3 de Transcrição de Octâmero/metabolismo , Adulto , Células-Tronco Adultas/metabolismo , Proteínas Argonautas/genética , Proteínas Argonautas/metabolismo , Diferenciação Celular , Proliferação de Células , Células Cultivadas , Proteínas de Ligação a DNA/genética , Regulação da Expressão Gênica no Desenvolvimento , Humanos , Família Multigênica , Regulação para CimaRESUMO
This study compared the marginal adaptation of direct composites under base materials with different elastic moduli. MOD cavities were prepared in 30 teeth. The cervical margin was placed 1 mm above the cementoenamel junction (CEJ) in one side and 1 mm below the CEJ in dentin in the other. The teeth were randomly divided into the following six groups (five teeth each) according to the base materials used: No base (Group 1), experimental flowable composite (Group 2), Helioflow (Ivoclar Vivadent) (Group 3), Tetric Flow (Group 4), Heliomolar HB (Ivoclar Vivadent) (Group 5) and Fuji II LC (Group 6). In Group 1, after etching the cavity enamel with 35% phosphoric acid, the cavities were primed and bonded with AdheSE, then filled with Tetric Ceram according to the manufacturer's instructions. In the other groups, after placing the base materials (1 mm thick) into the cavity, the cavity was filled with Tetric Ceram using the same methods as in Group 1. After storing the specimens in distilled water for seven days, they were finished and polished. Using stereomicroscopy at 150x magnification, marginal adaptation of the specimens was determined and the percentage of the imperfect margin (IM%) in the pre-loaded specimens was calculated. A mechanical load was applied using a custom-made Chewing simulator. All specimens were submitted to 600,000 load cycles at 49N with a frequency of 2Hz. The IM% in the post-load specimens was calculated. Repeated measured one-way ANOVA with Tukey was applied to compare the IM% in the six groups at the 95% confidence level. The results of statistical analysis indicated that the IM% was Group 3, 4, 6 < or = 2 < or = 5 < or = 1.
Assuntos
Resinas Compostas/química , Adaptação Marginal Dentária , Materiais Dentários/química , Restauração Dentária Permanente/métodos , Cimentos de Resina/química , Condicionamento Ácido do Dente , Resinas Acrílicas/química , Preparo da Cavidade Dentária/classificação , Esmalte Dentário/patologia , Polimento Dentário , Restauração Dentária Permanente/classificação , Dentina/patologia , Módulo de Elasticidade , Cimentos de Ionômeros de Vidro/química , Humanos , Teste de Materiais , Ácidos Fosfóricos/química , Resinas Sintéticas/química , Estresse Mecânico , Fatores de Tempo , Água/químicaRESUMO
An unsupervised competitive neural network for efficient clustering of Gaussian probability density function (GPDF) data of continuous density hidden Markov models (CDHMMs) is proposed in this paper. The proposed unsupervised competitive neural network, called the divergence-based centroid neural network (DCNN), employs the divergence measure as its distance measure and utilizes the statistical characteristics of observation densities in the HMM for speech recognition problems. While the conventional clustering algorithms used for the vector quantization (VQ) codebook design utilize only the mean values of the observation densities in the HMM, the proposed DCNN utilizes both the mean and the covariance values. When compared with other conventional unsupervised neural networks, the DCNN successfully allocates more code vectors to the regions where GPDF data are densely distributed while it allocates fewer code vectors to the regions where GPDF data are sparsely distributed. When applied to Korean monophone recognition problems as a tool to reduce the size of the codebook, the DCNN reduced the number of GPDFs used for code vectors by 65.3% while preserving recognition accuracy. Experimental results with a divergence-based k-means algorithm and a divergence-based self-organizing map algorithm are also presented in this paper for a performance comparison.