China
Africa
Explore chapters and articles related to this topic
Diversity, interconnectivity and sustainability
Published in Jan Bogg, Robert Geyer, Complexity, Science and Society, 2017
Peter M. Allen, Pierpaolo Andriani
The information distance matrix, in turn, can be used in hierarchical clustering, multidimensional scaling, or network graphs producing visualisations, which allow the reader to interpret the data in an intuitive manner. Event space thus becomes more a spatial tool than a spatial metaphor. Data clusters, and the distances between and within clusters, may be interpreted as indicators of innovativeness and imitation in the observed field of practice. Changes in data clustering in time sequences of data may be interpreted as trends. Isolated and persistent clusters may be interpreted as stable niches. The method was applied to a sampled corpus of practitioners’ self reports of attempts to implement ICT in education. By mapping the texts according to the calculated normalised information distances between them, diversity and complexity within and between samples is assessed and discussed, with regard to the context of regulation.
Multidimensional scaling
Published in Pat Dugard, John Todman, Harry Staines, Approaching Multivariate Analysis, 2010
Pat Dugard, John Todman, Harry Staines
As in cluster analysis, we can get SPSS to calculate a distance matrix from a data matrix of cases and variables, and the same considerations apply to the choice of distance measure. Alternatively we can enter a distance matrix as a lower or upper triangle like the one in Table 11.1. We note, in passing, that SPSS offers MDS of non-metric (ordinal) data as well as scaling of metric data (distances or proximities on an interval scale), though we will not be looking at any ordinal examples.
Morphometries of Craniofacial Form
Published in D. Dixon Andrew, A.N. Hoyte David, Ronning Olli, Fundamentals of Craniofacial Growth, 2017
Besides FEM, another comparatively recent coordinate-free method that needs to be mentioned is Euclidean Distance Matrix Analysis or EDMA (Lele, 1993; Lele and Richtsmeier, 1990, 1991, 1992; Richtsmeier et al., 1991). Recall that one of the properties of a desirable form representation is invariance to coordinate transformations such as uniform scaling, translation, rotation, and reflection (Section III). Such a representation is possible with EDMA (Lele, 1991). In brief, EDMA utilizes 3-D Cartesian coordinates of the homologous points to identify local areas of significant shape change (Corner and Richtsmeier, 1993). Three steps are involved: (1) the computation of a “mean form” from the distances of all possible pairs of landmarks (these individual distances, viewed as an individual Euclidean matrix representation, are then averaged to yield a mean matrix for the sample); (2) the calculation of a form difference matrix using a ratio of similarity between forms based on the individual pairwise distances between two mean forms being compared; and (3) this distance difference matrix is then sorted to identify the areas of maximum and minimum change. Local areas of maximum and minimum change are identified by how much the difference matrix values differ from one. These ratios do not distinguish between “size and shape,” as both aspects are included in the analysis. See Corner and Richtsmeier (1993) for a recent application of EDMA to the primate craniofacial complex. While this technique is independent of FEM, results are generally similar. This is perhaps not unexpected if both approaches are based on the same homologous point data set.
Functional distributional clustering using spatio-temporal data
Published in Journal of Applied Statistics, 2023
A. Venkatasubramaniam, L. Evers, P. Thakuriah, K. Ampountolas
Accordingly, let D be a distance matrix, where distance between cluster d. The CDFs corresponding to the clusters d at each subsequent iteration and this process continues until a single larger cluster containing every vertex in the network is obtained. In a hierarchical clustering approach, a partition then occurs at each iteration to determine non-overlapping clusters.
Phylogenetic analyses of 41 Y-STRs and machine learning-based haplogroup prediction in the Qingdao Han population from Shandong province, Eastern China
Published in Annals of Human Biology, 2023
Guang-Yao Fan, De-Zhi Jiang, Yao-Heng Jiang, Wei Song, Ying-Yun He, Nixon Austin Wuo
Allele and haplotype frequencies of Y-STR loci were determined by direct counting. Genetic diversity (GD) and haplotype diversity (HD) were calculated using Nei’s method (Nei and Tajima 1981). The discrimination capacity (DC) was equal to the total distinct haplotypes divided by the total number of haplotypes. Based on raw genotype data from the Qingdao Han and 51 other ethnic groups (Supplementary Table S1), DA genetic distances were calculated using the “Genet.dist” package under the R statistical environment (Takezaki and Nei 1996). The distance matrix was further transformed into a heatmap using the “ComplexHeatmap” package under the R statistical environment (Gu et al. 2016). A multidimensional scaling (MDS) plot was constructed using the IBM SPSS® software (Young 1970; Hansen 2005). In addition, a well-established machine learning (ML) technique and classification method, Linear Discriminant Analysis (LDA), was utilised for predicting categories (Venables et al. 2002). Also, under the R statistical environment, the LDA plot was created via the “lda” package to visualise how well it can separate the 24 Han populations. The multi-copy loci were excluded from the LDA. To better exhibit the phylogenetic relationship in neighbor-joining (N-J) trees, the DA distance matrices were also used to create the “newick” file using the PHYLIP v3.6.95 (Reynolds et al. 1983; Retief 2000) and Evolview v3 (Subramanian et al. 2019).
Telemedicine use for pediatric asthma care: a mixed methods study
Published in Journal of Asthma, 2022
Sarah C. Haynes,, Rory Kamerman-Kretzmer,, Shahabal S. Khan,, Stephanie Crossen,, Monica K. Lieng,, James P. Marcin,, Nicholas J. Kenyon,, Christopher H. Kim,
We abstracted data from the UC Davis EHR on all patients ages 2–24 who had any visit for a primary diagnosis of asthma during the six months following California’s statewide shelter-in-place order (March 19, 2020–September 30, 2020). We chose to include young adults (ages 18–24) in our analysis of pediatric asthma care because many of these patients are still seen in pediatric pulmonology clinics as they transition to independent care. Our outcome of interest was telemedicine use for healthcare encounters with a primary diagnosis of asthma. Data abstracted included the visit type (in-person, telemedicine, or phone), clinical department, and patient sex, age, race/ethnicity, primary language, insurance, and zip code. We matched patient zip codes to the Federal Office of Rural Health Policy (FORHP) list of eligible zip codes to identify rural/urban residence (9). We calculated driving distance from the patient’s home to the UC Davis Health ambulatory clinic using the gmapsdistance package in R (version 3.6.1) to connect to the Google Maps Distance Matrix Application Programming Interface (10,11). As this interface requires using an estimated future time to do the calculations, we used July 7th, 2020 10:00AM to estimate driving times during the day without the influence of morning rush hour. These calculations were completed in February 2020, before the California shelter-in-place affected traffic patterns.