global accuracy
Recently Published Documents





Xueying Yu ◽  
Yanlin Shao ◽  
David R. Fuhrman

Abstract It is essential for a Navier-Stokes equations solver based on a projection method to be able to solve the resulting Poisson equation accurately and efficiently. In this paper, we present numerical solutions of the 2D Navier-Stokes equations using the fourth-order generalized harmonic polynomial cell (GHPC) method as the Poisson equation solver. Particular focus is on the local and global accuracy of the GHPC method on non-uniform grids. Our study reveals that the GHPC method enables use of more stretched grids than the original HPC method. Compared with a second-order central finite difference method (FDM), global accuracy analysis also demonstrates the advantage of applying the GHPC method on stretched non-uniform grids. An immersed boundary method is used to deal with general geometries involving the fluid-structure-interaction problems. The Taylor-Green vortex and flow around a smooth circular cylinder and square are studied for the purpose of verification and validation. Good agreement with reference results in the literature confirms the accuracy and efficiency of the new 2D Navier-Stokes equation solver based on the present immersed-boundary GHPC method utilizing non-uniform grids. The present Navier-Stokes equations solver uses second-order FDM for the discretization of the diffusion and advection terms, which may be replaced by other higher-order schemes to further improve the accuracy.

2022 ◽  
Vol 15 (1) ◽  
pp. 35
Shekar Shetty ◽  
Mohamed Musa ◽  
Xavier Brédart

In this study, we apply several advanced machine learning techniques including extreme gradient boosting (XGBoost), support vector machine (SVM), and a deep neural network to predict bankruptcy using easily obtainable financial data of 3728 Belgian Small and Medium Enterprises (SME’s) during the period 2002–2012. Using the above-mentioned machine learning techniques, we predict bankruptcies with a global accuracy of 82–83% using only three easily obtainable financial ratios: the return on assets, the current ratio, and the solvency ratio. While the prediction accuracy is similar to several previous models in the literature, our model is very simple to implement and represents an accurate and user-friendly tool to discriminate between bankrupt and non-bankrupt firms.

2021 ◽  
Vol 11 (1) ◽  
Rafael Mamede ◽  
Florbela Pereira ◽  
João Aires-de-Sousa

AbstractMachine learning (ML) algorithms were explored for the classification of the UV–Vis absorption spectrum of organic molecules based on molecular descriptors and fingerprints generated from 2D chemical structures. Training and test data (~ 75 k molecules and associated UV–Vis data) were assembled from a database with lists of experimental absorption maxima. They were labeled with positive class (related to photoreactive potential) if an absorption maximum is reported in the range between 290 and 700 nm (UV/Vis) with molar extinction coefficient (MEC) above 1000 Lmol−1 cm−1, and as negative if no such a peak is in the list. Random forests were selected among several algorithms. The models were validated with two external test sets comprising 998 organic molecules, obtaining a global accuracy up to 0.89, sensitivity of 0.90 and specificity of 0.88. The ML output (UV–Vis spectrum class) was explored as a predictor of the 3T3 NRU phototoxicity in vitro assay for a set of 43 molecules. Comparable results were observed with the classification directly based on experimental UV–Vis data in the same format.

2021 ◽  
Vol 922 (2) ◽  
pp. 204
John F. Suárez-Pérez ◽  
Yeimy Camargo ◽  
Xiao-Dong Li ◽  
Jaime E. Forero-Romero

Abstract Precise cosmic web classification of observed galaxies in massive spectroscopic surveys can be either highly uncertain or computationally expensive. As an alternative, we explore a fast Machine Learning-based approach to infer the underlying dark matter tidal cosmic web environment of a galaxy distribution from its β-skeleton graph. We develop and test our methodology using the cosmological magnetohydrodynamic simulation Illustris-TNG at z = 0. We explore three different tree-based machine-learning algorithms to find that a random forest classifier can best use graph-based features to classify a galaxy as belonging to a peak, filament, or sheet as defined by the T-Web classification algorithm. The best match between the galaxies and the dark matter T-Web corresponds to a density field smoothed over scales of 2 Mpc, a threshold over the eigenvalues of the dimensionless tidal tensor of λ th = 0.0, and galaxy number densities around 8 × 10−3 Mpc−3. This methodology results on a weighted F1 score of 0.728 and a global accuracy of 74%. More extensive tests that take into account light-cone effects and redshift space distortions are left for future work. We make one of our highest ranking random forest models available on a public repository for future reference and reuse.

G. Hao ◽  
A. Ni ◽  
Y.J. Chang ◽  
K. Hall ◽  
S.H. Lee ◽  

BACKGROUND: Currently there is limited information to guide health professionals regarding the optimal time frame to initiate safe and effective oral feedings to preterm infants. The study aims to revise and validate a streamlined version of the “Traditional Chinese-Preterm Oral Feeding Readiness Assessment Scale”, the TC-POFRAS®, and evaluate its construct validity in the clinical decisions regarding feeding readiness of preterm infants. METHODS: Eighty-one clinically stable preterm infants were assessed using the TC-POFRAS for oral feeding readiness. Item-total correlation analysis was used to check if any item was inconsistent with the averaged TC-POFRAS scores. Cronbach’s α coefficient was used to evaluate the inter-item consistency. Exploratory factor analysis was used to determine the coherence of variables to reorganize assessment domains. The revised version of TC-POFRAS (TC-POFRAS®) was developed and a new cut-off score based on discriminant accuracy was established. RESULTS: Based on the results from statistical analysis, five items (“lips posture,” “tongue posture,” “biting reflex,” “gag reflex,” and “tongue cupping”) were deleted from the original TC-POFRAS to form the TC-POFRAS®. The TC-POFRAS®’s global accuracy was 92.1%. The cut-off value of 19 was the one that presented the most optimization of sensitivity based on specificity. The TC-POFRAS® was reconstructed into corrected gestational age and five behavioral domains. CONCLUSIONS: The TC-POFRAS® is considered a valid, safe, and accurate objective instrument to assist health professionals to initiate oral feeding of the preterm infants.

2021 ◽  
Vol 7 (20) ◽  
pp. 202128
Antonia Sueli Silva Sousa ◽  
Paulo Roberto Mendes Pereira ◽  
Audivan Ribeiro Garcês Júnior

QUALITY ASSESSMENT OF LANDSAT 8 IMAGE CLASSIFIERS IN A SAGA GIS COMPUTER ENVIRONMENT FOR LAND COVERING MAPPING IN THE CERRADO BIOMEEVALUACIÓN DE LA CALIDAD DE LOS CLASIFICADORES DE IMAGEN LANDSAT 8 EN UN ENTORNO COMPUTACIONAL SAGA GIS PARA EL MAPEO DE COBERTURA DE TIERRAS EN EL BIOMA DE CERRADORESUMOUma das principais aplicações das imagens de satélites é a caracterização da cobertura terrestre, que a partir do uso de técnicas de classificação permite monitorar as transformações espaciais da superfície terrestre. O Sistema Automatizado de Análise Geociêntífica – Saga Gis apresenta um conjunto de ferramentas voltado à análise geográfica, incluindo pacotes de classificação de imagens digitais, onde se destacam os classificadores: Maxver, Mahalanobis, distância mínima, paralelepípedo. O objetivo deste artigo é avaliar o potencial dos classificadores de imagens do Saga Gis no bioma Cerrado, sendo objeto de estudo, o município de Brejo-MA. Foi utilizada uma imagem Landsat 8 de 2017, com resolução espacial de 30 metros. A metodologia consistiu na aplicação de um conjunto de técnicas de tratamento digital de imagens, segmentação, extração de atributos e classificação. A análise dos dados pautou-se na comparação visual e análise da exatidão global e de índice Kappa. O classificador Maxver apresentou os melhores resultados para o Kappa e exatidão global, já os piores valores foram associados ao classificador paralelepípedo.Palavras-chave: Geotecnologia; Processamento de Imagem; Acurácia, Mapeamento. ABSTRACTOne of the main applications of satellite images is the characterization of terrestrial coverage, which from the use of classification techniques allows to monitor the spatial transformations of the terrestrial surface. The System for Automated Geoscientific Analyzes-Saga Gis presents a set of tools aimed at geographic analysis, including digital image classification packages, in which the classifiers stand out: Maxver, Mahalanobis, minimum distance, parallelepiped. The objective of this article is to evaluate the potential of the Saga Gis image classifiers in the Cerrado biome, being the object of study, the municipality of Brejo-MA. It was to use a Landsat 8 image (2017), with a spatial resolution of 30 meters. The methodology consisted of applying a set of techniques for digital image processing, segmentation, attribute extraction and classification. Data analysis was based on visual comparison and analysis of global accuracy and Kappa index. The Maxver classifier presented the best results for Kappa and overall accuracy, whereas the worst values were associated with the parallelepiped classifier.Keywords: Geotechnology; Image Processing; Accuracy; Mapping.RESUMENUna de las principales aplicaciones de las imágenes de satélite es la caracterización de la cobertura terrestre, que, a partir del uso de técnicas de clasificación, permite el seguimiento de las transformaciones espaciales de la superficie terrestre. El Sistema de Análisis Geocientífico Automatizado (Saga Gis) presenta un conjunto de herramientas orientadas al análisis geográfico, que incluyen paquetes de clasificación de imágenes digitales, en los que destacan los clasificadores: Maxver, Mahalanobis, distancia mínima, paralelepípedo. El objetivo de este artículo es evaluar el potencial de los clasificadores de imágenes Saga Gis en el bioma del Cerrado, siendo objeto de estudio, el municipio de Brejo-MA. Se utilizó una imagen Landsat 8 de 2017 con una resolución espacial de 30 metros. La metodología consistió en aplicar un conjunto de técnicas de procesamiento, segmentación, extracción de atributos y clasificación de imágenes digitales. El análisis de los datos se basó en la comparación visual y el análisis de la precisión global y el índice Kappa. El clasificador Maxver presentó los mejores resultados para Kappa y precisión general, mientras que los peores valores se asociaron con el clasificador paralelepípedo.Palabras clave: Geotecnología; Procesamiento de imágenes; Precisión; Mapeo.

2021 ◽  
Rebecca White ◽  
Ronald Swatzyna

Abstract A new pre-processing approach of EEG data to detect topological EEG features has been applied to a continuous segment of artifact-free EEG data lasting 10 minutes in ASCII format derived from 50 ASD children and 50 children with other Neuro-psychiatric disorders, matched for age and male/female ratios. Each EEG was manipulated using a Cin-Cin algorithm, based on an input vector characterized by a linear composition of city-block matrix distances among19 electrodes. From the resulting triangular matrix of 171 numbers expressing all of the one-by-one distances among the 19 electrodes a minimum spanning tree(MST) is calculated. Electrode identification serial codes, sorted according to the decreasing number of links in MST, and the number of links in MST are taken as input vectors for machine learning systems. With this method all the content of an EEG is transformed in 38 numbers which represent the input vectors for machine learning systems classifiers. Machine learning systems have been applied to build up a predictive model to distinguish between the two diagnostic classes. The best machine learning system (KNN algorithm) obtained a global accuracy of 93.2% (92.37 % sensitivity and 94.03 % specificity) in differentiating ASD subjects from NPD subjects. In conclusion the results obtained in this study suggest that the two new pre-processing methods introduced, in particular the MST algorithm, have great potential to allow a machine learning system to discriminate EEGs obtained from subjects with autism from EEGs obtained from subjects affected by other psychiatric disorders.

2021 ◽  

The greatest challenge in creating digital material twins and FE mesh from μCT images of composite reinforcements is the lack of a robust and versatile tool for training μCT images. Here, we have used deep convolutional neural networks (DCNN) for segmenting μCT images of a multi-layer plain-weave fiber reinforcement. A set of raw 2D image slices extracted from the gray-scale volume of a single-layer reinforcement was used to train a DCNN using manually annotated images. The trained network was tested against the manually segmented ground truth images and it performed exceptionally well with a global accuracy of more than 96%. The trained DCNN was then used to segment unseen images from a multilayer stack of the fabric with good accuracy. The work presented here provides a robust and efficient framework of segmenting CT scan images of fiber reinforcements for generating digital material twins and FE mesh of fiber reinforcements.

2021 ◽  
pp. 136216882110434
Simona Floare Bora

This article intends to add to the rising discussion related to the employment of authentic plays and drama within a high school compulsory curriculum for enhancing learners’ foreign language (L2) oral skills. In particular, it examines the pedagogical use of authentic contemporary plays for developing learners’ L2 oral production in terms of (1) complexity – syntactic and mean length of AS-units (MLAS) and (2) accuracy – global and pronunciation accuracy. For this purpose, a class of 10 final year high school students with a lower-intermediate to upper-intermediate level of language in an Italian context was exposed longitudinally to a blended-drama approach – the use of literary play scripts, drama games and techniques, and a full-scale performance – conducted over two terms for a total of 40 hours in-class lessons. A control group was taught through a traditional approach over the same period. Quantitative data were collected through a pre-test/post-test design with three tasks under different conditions regarding status and interaction: oral proficiency interview (OPI), story-retelling and guided role-play (GRP). Findings revealed that drama significantly improved learners’ pronunciation accuracy, syntactic complexity and MLAS. There was no significant statistical result on global accuracy between the two groups. Pedagogical implications for teaching practice will be discussed.

2021 ◽  
Vol 7 (9) ◽  
pp. 180
Ana C. Morgado ◽  
Catarina Andrade ◽  
Luís F. Teixeira ◽  
Maria João M. Vasconcelos

With the increasing adoption of teledermatology, there is a need to improve the automatic organization of medical records, being dermatological image modality a key filter in this process. Although there has been considerable effort in the classification of medical imaging modalities, this has not been in the field of dermatology. Moreover, as various devices are used in teledermatological consultations, image acquisition conditions may differ. In this work, two models (VGG-16 and MobileNetV2) were used to classify dermatological images from the Portuguese National Health System according to their modality. Afterwards, four incremental learning strategies were applied to these models, namely naive, elastic weight consolidation, averaged gradient episodic memory, and experience replay, enabling their adaptation to new conditions while preserving previously acquired knowledge. The evaluation considered catastrophic forgetting, accuracy, and computational cost. The MobileNetV2 trained with the experience replay strategy, with 500 images in memory, achieved a global accuracy of 86.04% with only 0.0344 of forgetting, which is 6.98% less than the second-best strategy. Regarding efficiency, this strategy took 56 s per epoch longer than the baseline and required, on average, 4554 megabytes of RAM during training. Promising results were achieved, proving the effectiveness of the proposed approach.

Sign in / Sign up

Export Citation Format

Share Document