Introduction Multiple definitions for “difficult to treat” patients (DTP) were given throughout the years. While most authors focus on diagnoses, others focus on clinical, social and demographic ...factors, which should be regarded as factors of bad prognosis and elevated costs for the healthcare systems. Objectives To identify and haracterize DTP patients admitted in acute ward, based on practical criteria. Methods Through the hospital’s IT services, all acute inpatient episodes at Centro Hospitalar Psiquiátrico de Lisboa were collected, since 2017. Cluster analysis was performed, regarding number of previous admissions (PA) and days of admission. Descriptive and comparative statistics (with multiple comparisons) for the different clusters, regarding age, gender, diagnosis at discharge (according to ICD10), and, to the DTP, previous medical following, compliance to medication, and substance use at admission. Results Three clusters were identified: (C1, n=5861) a larger, uncharacteristic one; (C2, n=1168) with a higher number of PA (average of 8, versus less than 2 on the others); and (C3, n=1462) with higher number of days of admissions (58 versus less than 16). Statistical significance was found regarding age (higher in C3), gender (more men in C2), nationality (C1 with more foreigners). Regarding diagnosis at discharge, statistical difference was found between the 3 groups: C1 has significantly less patients with Schizophrenia (11% versus 30% in the others), but more depressive (21% versus 6% in C2 and 12% in C3) and neurotic disorders. C2 presented less dementias (0,5% versus 3% in C1 and 10% in C3) and delusional disorders, but more bipolar disorders (24% versus 15% in C1 and C3); C3 represented less episodes due to substance abuse (alcohol or others) and personality disorders. In both C2 and C3, no psychiatric consultation happened in the 3 months prior admission to around 40% of episodes, and 50% had stopped medication. The majority had only oral medication. Almost 24% of C2 tested positive for cannabinoids, with no differences regarding other substances. Conclusions These findings allow the definition of 2 kinds of DTP, which present unique characteristics but some common features (namely poor adherence to consultations and are in therapeutic compliance). An assertive multidisciplinary approach, focused on current treatment and relapse prevention (including social structures, more frequent clinical follow-up, and rehabilitation centers), will be the key to their treatment. Disclosure of Interest None Declared
The heterogeneity of cognitive profiles among psychiatric patients has been reported to carry significant clinical information. However, how to best characterize such cognitive heterogeneity is still ...a matter of debate. Despite being well suited for clinical data, cluster analysis techniques, like the Two-Step and the Latent Class, received little to no attention in the literature. The present study aimed to test the validity of the cluster solutions obtained with Two-Step and Latent Class cluster analysis on the cognitive profile of a cross-diagnostic sample of 387 psychiatric inpatients. Two-Step and Latent Class cluster analysis produced similar and reliable solutions. The overall results reported that it is possible to group all psychiatric inpatients into Low and High Cognitive Profiles, with a higher degree of cognitive heterogeneity in schizophrenia and bipolar disorder patients than in depressive disorders and personality disorder patients.
This book focuses on a few of the most important clustering algorithms, providing a detailed account of these major models in an information retrieval context. The beginning chapters introduce the ...classic algorithms in detail, while the later chapters describe clustering through divergences and show recent research for more advanced audiences.
Los últimos años de la escuela secundária se caracterizan por desafíos interpersonales y académicos, además de altos niveles de deserción y fracaso escolar. Sin embargo, son pocos los estudios que ...tienen como objetivo investigar este período de la trayectoria educativa, con el fin de identificar recursos personales y contextuales. El objetivo fue analizar perfiles de ajuste psicosocial de los estudiantes, considerando factores de riesgo (exposición a la violencia y discriminación diaria), protección (apoyo social y clima escolar) y ajuste indicador (satisfacción con la vida). Participaron 709 estudiantes que cursaban los grados 7º, 8º y 9º de escuelas públicas. Análisis de cluster identificaron perfiles: resiliente, con valores altos de indicadores de riesgo con buen ajuste; vulnerable, con índices de alto riesgo y bajo ajuste. Se concluye que invertir en la reducción de factores de riesgo y potenciación de factores protectores, a través de programas preventivos, es fundamental para el desarrollo.
We present a review of second language researchers’ use of cluster analysis, an advanced statistical method still uncommon but increasingly used to identify groups or patterns in a dataset and to ...examine group differences. After describing key methodological considerations in conducting cluster analysis, we present a methodological synthesis of 65 studies published between 1989 and 2018 that employed cluster analysis. We specifically review the use of cluster analysis for themes of usage and reporting practices. Our findings indicate that hierarchical cluster analysis and K‐means cluster analysis were the most commonly used cluster methods, but the widespread use of these two methods tended not to be accompanied by sound reporting practices, particularly when justifying cluster solutions. In our analysis, we highlight concerns related to reporting and evaluation. For future use and to inform methodological practices in second language research, we briefly report on a sample study of cluster analysis that uses open data.
Cluster analysis of IRPJ precedents in CARF Costa, Fabiano de Castro Liberato; Martinez, Antonio Lopo; Klann, Roberto Carlos
Revista de contabilidade e organizações,
01/2023, Letnik:
17
Journal Article
Recenzirano
Odprti dostop
O objetivo deste estado foi agrupar acordaos do Conselho Administrativo de Recursos Piscáis i'(ART) relacionados ao Imposto de Renda Pessoa Jurídica (IRPJ), prolatados entre 2016 e 2020, empregando ...técnicas de aprendizado de máquina (VIL) para a clusterizaçao de documentos textuais. A análise resultan em 13 clusters exclusivos, um ochado inédito na literatura contabil tributaria no Brasil. Essa identificaçao é relevante para o (ARI', contribuintes, administraçao tributaria e profissionais contábeis e tributaristas envolvidos em questöes contábeis e tributarias relacionadas ao IRPJ. Os algoritmos de AÍL utilizados mostraram-se eficientes na resoluçao de problemas complexos de processamento de linguagem natural (PLN), como criar representaçöes vetoriais de temos e identificar temáticas em dados nao estmturados, fomecendo contribuiçöes valiosas para o entendimento de materias controversas no IRPJ á luz da jurisprudencia administrativa. A clusterizaçao de precedentes se traduz em maior acessibilidade e análise de padrees nos jidgamentos, facilitando a tomada de decisöes na contabilidade tributaria.
Abstract
Summary
VarSelLCM allows a full model selection (detection of the relevant features for clustering and selection of the number of clusters) in model-based clustering, according to classical ...information criteria. Data to be analyzed can be composed of continuous, integer and/or categorical features. Moreover, missing values are managed, without any pre-processing, by the model used to cluster with the assumption that values are missing completely at random. Thus, VarSelLCM also allows data imputation by using mixture models. A Shiny application is implemented to easily interpret the clustering results.
Availability and implementation
VarSelLCM is available to download at https://CRAN.R-project.org/package=VarSelLCM/.
Tutorial
vignette is available online at http://varsellcm.r-forge.r-project.org/
Supplementary information
Supplementary data are available at Bioinformatics online.
Display omitted
•A novel criterion for clustering ore samples for liberation studies is proposed.•It is easily obtained from the liberation spectrum for a top size of 1 mm.•The criterion has been ...experimentally validated for iron ore samples.•It is potentially applicable to other types of ore.•A clustering approach significantly reduces laboratory effort.
Correctly determining the optimal grinding size is crucial for mineral processors, yet it poses several challenges. Still popular among practitioners, the empirical determination of the grinding target involves time-consuming and resource-intensive laboratory routines employed for generating quality, recovery and energy curves from which the target size is set. This approach becomes even more challenging for mines with diverse ore textures, requiring extensive testing on numerous samples to accurately capture the behaviour of the entire ore body. Consequently, a more efficient experimental method is highly desirable. A solution for reducing the laboratory effort is clustering samples according to the mineral liberation characteristics. This paper introduces a novel multivariate criterion for statistical clustering of ore samples for mineral liberation studies, derived from SEM-based automated mineralogy data. The criterion considers three variables: two coefficients (named A and k) derived from the exponential correlation equation between degree of liberation and the top size of the size fraction, and the overall degree of liberation, both easily obtained from the liberation spectrum for a top size of 1 mm. The effectiveness of the proposed criterion has been experimentally demonstrated. It was concluded that the clustering process correctly grouped samples with similar mineral liberation characteristics. The clustering approach can be used to significantly reduce the laboratory effort on liberation studies involving a great number of samples of different types of ore. Although developed using iron ore samples, the criterion's theoretical basis suggests its potential for adaptation to other types of ore.
Abstract
Introduction
Spindles are currently defined clinically based on observed patterns in the EEG waveform trace, with automated methods seeking to replicate visual scoring by experts. Recent ...work suggests that sleep spindles may be more readily observed as time-frequency peaks in the EEG spectrogram. This study compares spectral peaks in the multitaper spectrogram to expert and automatic detection scoring, characterizes the variability of spindles across a night, and investigates topographical and temporal clustering of spindles within individual EEG records.
Methods
We compared spectral peaks, expert scoring, and automatic detection in two datasets (DREAMS, and a high-density control study). Peaks were identified using multitaper spectral estimation and the peak prominence of the normalized power spectrum for each channel. Spatiotemporal variability analysis was performed using cluster and pattern recognition algorithms including penalized sorting of channel activation order, 2D-cross correlation, PCA and UMAP cluster analysis, and the seqNMF method.
Results
Spectral peaks were shown to be highly robust to and easily differentiated from broadband noise, occuring at rates (10-16 per min) far exceeding spindle rates reported in literature (~2.5 per min). Expert scoring and automated scoring failed to capture clear spectral peaks in the time-frequency domain, indicating an underreporting of the phenomenology. No apparent clustering or patterns of sleep spindle-like activity was observed using the proposed methods, suggesting high variability of spatiotemporal evolution of spindles.
Conclusion
These results suggest that the difficulty of time-domain visual scoring of spindles causes an artificially low estimate of the underlying phenomenology, which is mirrored in the assumptions implicit in the thresholds of automated scorers. This work shows that spindles are highly variable in their spatiotemporal evolution, suggesting that there is no optimal single electrode for analysis and casting doubt on the presence of a single cortical generation mechanism. We must therefore revisit the concept of the spindle using the time-frequency domain to more robustly characterize underlying phenomenology.
Support
National Institute Of Neurological Disorders And Stroke Grant R01 NS-096177
PDF
