Your browser doesn't support javascript.
Mostrar: 20 | 50 | 100
Resultados 1 - 20 de 5.195
Filtrar
1.
Water Sci Technol ; 80(3): 466-477, 2019 Aug.
Artigo em Inglês | MEDLINE | ID: mdl-31596258

RESUMO

Wetlands are among the most productive ecosystems that provide services ranging from flood control to climate change mitigation. Wetlands are also critical habitats for the survival of numerous plant and animal species. In this study, we used satellite remote sensing techniques for classification and change detection at an internationally important wetland (Ramsar Site) in Turkey. Sultan Marshes is located at the center of semi-arid Develi closed basin. The wetlands have undergone significant changes since the 1980s due to changes in water flow regimes, but changes in recent years have not been sufficiently explored yet. In this study, we focused on the changes from 2005 to 2012. Two multispectral ASTER images with spatial resolution of 15 m, acquired on June 11, 2005 and May 20, 2012, were used in the analyses. After geometric correction, the images were classified into four information classes, namely water, marsh, agriculture, and steppe. The applicability of three classification methods (i.e. maximum likelihood (MLH), multi-layer perceptron type artificial neural networks (ANN) and support vector machines (SVM)) was assessed. The differences in classification accuracies were evaluated by the McNemar's test. The changes in the Sultan Marshes were determined by the post classification comparison method using the most accurate classified images. The results showed that the highest overall accuracy in image classifications was achieved with the SVM method. It was observed that marshes and steppe areas decreased while water and agricultural areas expanded from 2005 to 2012. These changes could be the results of water transfers to the marshes from neighboring watershed.


Assuntos
Monitoramento Ambiental/métodos , Máquina de Vetores de Suporte , Áreas Alagadas , Conservação dos Recursos Naturais , Ecossistema , Eugenol , Redes Neurais (Computação) , Turquia , Óxido de Zinco
2.
Artigo em Chinês | MEDLINE | ID: mdl-31594134

RESUMO

Objective: To establish a CT image radiomics-based prediction model for the differential diagnosis of silicosis and tuberculosis nodules. Methods: A total of 53 patients with silicosis and 89 patients with tuberculosis who underwent routine CT scans in Suzhou Fifth People's Hospital from January to August, 2018 were enrolled in this study. AK/ITK software was used to segment the images to obtain 139 silicosis lesions and 119 tuberculosis lesions. For each lesion image, 396 features were extracted, and feature dimension reduction was applied to select the most characteristic feature subset. Support vector machine (SVM) , feedforward back propagation neural network (FNN-BP) , and random forest (RF) were implemented using R software (Rstudio V1.1.463) , and the algorithm that achieved the largest area under of the receiver operating characteristic (ROC) curve (AUC) was selected as the final prediction model. Results: RF was the best prediction model for the differential diagnosis of silicosis and tuberculosis nodules, with an accuracy of 83.1%, a sensitivity of 0.76, a specificity of 0.9, and an AUC of 0.917 (95% confidence interval: 0.8431-0.9758) . RF had a significantly larger AUC than SVM and FNN-BP (P<0.05) . Conclusion: CT image-based RF prediction model can be used to differentially diagnose silicosis and tuberculosis nodules.


Assuntos
Silicose/diagnóstico por imagem , Tuberculose/diagnóstico por imagem , Diagnóstico Diferencial , Humanos , Interpretação de Imagem Assistida por Computador , Modelos Teóricos , Redes Neurais (Computação) , Curva ROC , Máquina de Vetores de Suporte , Tomografia Computadorizada por Raios X
3.
Sheng Wu Yi Xue Gong Cheng Xue Za Zhi ; 36(5): 755-762, 2019 Oct 25.
Artigo em Chinês | MEDLINE | ID: mdl-31631623

RESUMO

Autoimmune pancreatitis (AIP) is a unique subtype of chronic pancreatitis, which shares many clinical presentations with pancreatic ductal adenocarcinoma (PDA). The misdiagnosis of AIP often leads to unnecessary pancreatic resection. 18F-FDG positron emission tomography/ computed tomography (PET/CT) could provide comprehensive information on the morphology, density, and functional metabolism of the pancreas at the same time. It has been proved to be a promising modality for noninvasive differentiation between AIP and PDA. However, there is a lack of clinical analysis of PET/CT image texture features. Difficulty still remains in differentiating AIP and PDA based on commonly used diagnostic methods. Therefore, this paper studied the differentiation of AIP and PDA based on multi-modality texture features. We utilized multiple feature extraction algorithms to extract the texture features from CT and PET images at first. Then, the Fisher criterion and sequence forward floating selection algorithm (SFFS) combined with support vector machine (SVM) was employed to select the optimal multi-modality feature subset. Finally, the SVM classifier was used to differentiate AIP from PDA. The results prove that texture analysis of lesions helps to achieve accurate differentiation of AIP and PDA.


Assuntos
Adenocarcinoma/diagnóstico por imagem , Doenças Autoimunes/diagnóstico por imagem , Neoplasias Pancreáticas/diagnóstico por imagem , Pancreatite/diagnóstico por imagem , Algoritmos , Diagnóstico Diferencial , Fluordesoxiglucose F18 , Humanos , Tomografia Computadorizada com Tomografia por Emissão de Pósitrons , Máquina de Vetores de Suporte
4.
Br J Anaesth ; 123(5): 688-695, 2019 11.
Artigo em Inglês | MEDLINE | ID: mdl-31558311

RESUMO

BACKGROUND: Postoperative mortality occurs in 1-2% of patients undergoing major inpatient surgery. The currently available prediction tools using summaries of intraoperative data are limited by their inability to reflect shifting risk associated with intraoperative physiological perturbations. We sought to compare similar benchmarks to a deep-learning algorithm predicting postoperative 30-day mortality. METHODS: We constructed a multipath convolutional neural network model using patient characteristics, co-morbid conditions, preoperative laboratory values, and intraoperative numerical data from patients undergoing surgery with tracheal intubation at a single medical centre. Data for 60 min prior to a randomly selected time point were utilised. Model performance was compared with a deep neural network, a random forest, a support vector machine, and a logistic regression using predetermined summary statistics of intraoperative data. RESULTS: Of 95 907 patients, 941 (1%) died within 30 days. The multipath convolutional neural network predicted postoperative 30-day mortality with an area under the receiver operating characteristic curve of 0.867 (95% confidence interval [CI]: 0.835-0.899). This was higher than that for the deep neural network (0.825; 95% CI: 0.790-0.860), random forest (0.848; 95% CI: 0.815-0.882), support vector machine (0.836; 95% CI: 0.802-870), and logistic regression (0.837; 95% CI: 0.803-0.871). CONCLUSIONS: A deep-learning time-series model improves prediction compared with models with simple summaries of intraoperative data. We have created a model that can be used in real time to detect dynamic changes in a patient's risk for postoperative mortality.


Assuntos
Aprendizado Profundo , Complicações Pós-Operatórias/mortalidade , Procedimentos Cirúrgicos Operatórios/mortalidade , Algoritmos , Comorbidade , Humanos , Missouri/epidemiologia , Redes Neurais (Computação) , Período Pós-Operatório , Valor Preditivo dos Testes , Estudos Retrospectivos , Medição de Risco/métodos , Máquina de Vetores de Suporte
5.
Waste Manag ; 100: 10-17, 2019 Dec.
Artigo em Inglês | MEDLINE | ID: mdl-31493684

RESUMO

In recent years, there has been a significant increase in the number of end-of-life vehicles (ELVs) in China. The traditional methods that rely primarily on manual sorting are hard to meet the requirements anymore. To solve the low intelligence and efficiency of separating non-ferrous metals, a machine vision based system was made to separate non-ferrous metals from ELVs, and the influences of the classification algorithm and operation parameters on the separation efficiency of the system were investigated. With the use of a principle component analysis/support vector machine (PCA-SVM) algorithm and decrease the number of features to three, the achieved recognition accuracy was 96.64%, and the computational speed was sufficiently high. Response surface methodology and FLUENT numerical simulation were employed to study the influence of operation parameters by evaluating the separation distance between copper and aluminum. The results indicated that the separation distance decreased in accordance with an increase in the speed of the conveyor belt (v), and increased in accordance with an increase in the air pressure of the nozzle (P) and separation height (H). With an increase in the angle of nozzle (α), there was a decrease in the separation distance after an initial increase, and the maximum value was reached at a nozzle angle 40°. The optimal operation parameters in this study were v = 1.4 m/s, P = 0.6 MPa, H = 0.6 m, α = 40°. The separation accuracy and purity of the system were greater than 85% using the proposed optimal classification algorithm and abovementioned operation parameters.


Assuntos
Algoritmos , Máquina de Vetores de Suporte , Alumínio , China , Análise de Componente Principal
6.
Sheng Wu Yi Xue Gong Cheng Xue Za Zhi ; 36(4): 521-530, 2019 Aug 25.
Artigo em Chinês | MEDLINE | ID: mdl-31441251

RESUMO

Atrial fibrillation (AF) is one of the most common arrhythmias, which does great harm to patients. Effective methods were urgently required to prevent the recurrence of AF. Four methods were used to analyze RR sequence in this paper, and differences between Pre-AF (preceding an episode of AF) and Normal period (far away from episodes of AF) were analyzed to find discriminative criterion. These methods are: power spectral analysis, approximate entropy (ApEn) and sample entropy (SpEn) analysis, recurrence analysis and time series symbolization. The RR sequence data used in this research were downloaded from the Paroxysmal Atrial Fibrillation Prediction Database. Supporting vector machine (SVM) classification was used to evaluate the methods by calculating sensitivity, specificity and accuracy rate. The results showed that the comprehensive utilization of recurrence analysis parameters reached the highest accuracy rate (95%); power spectrum analysis took second place (90%); while the results of entropy analyses and time sequence symbolization were not satisfactory, whose accuracy were both only 70%. In conclusion, the recurrence analysis and power spectrum could be adopted to evaluate the atrial chaotic state effectively, thus having certain reference value for prediction of AF recurrence.


Assuntos
Fibrilação Atrial/diagnóstico , Entropia , Átrios do Coração/fisiopatologia , Humanos , Recidiva , Sensibilidade e Especificidade , Máquina de Vetores de Suporte
7.
Sheng Wu Yi Xue Gong Cheng Xue Za Zhi ; 36(4): 531-540, 2019 Aug 25.
Artigo em Chinês | MEDLINE | ID: mdl-31441252

RESUMO

Individual differences of P300 potentials lead to that a large amount of training data must be collected to construct pattern recognition models in P300-based brain-computer interface system, which may cause subjects' fatigue and degrade the system performance. TrAdaBoost is a method that transfers the knowledge from source area to target area, which improves learning effect in the target area. Our research purposed a TrAdaBoost-based linear discriminant analysis and a TrAdaBoost-based support vector machine to recognize the P300 potentials across multiple subjects. This method first trains two kinds of classifiers separately by using the data deriving from a small amount of data from same subject and a large amount of data from different subjects. Then it combines all the classifiers with different weights. Compared with traditional training methods that use only a small amount of data from same subject or mixed different subjects' data to directly train, our algorithm improved the accuracies by 19.56% and 22.25% respectively, and improved the information transfer rate of 14.69 bits/min and 15.76 bits/min respectively. The results indicate that the TrAdaBoost-based method has the potential to enhance the generalization ability of brain-computer interface on the individual differences.


Assuntos
Interfaces Cérebro-Computador , Potencial Evocado P300 , Máquina de Vetores de Suporte , Algoritmos , Análise Discriminante , Eletroencefalografia , Humanos
8.
BMC Bioinformatics ; 20(1): 415, 2019 Aug 06.
Artigo em Inglês | MEDLINE | ID: mdl-31387547

RESUMO

BACKGROUND: Predicting the effect of drug-drug interactions (DDIs) precisely is important for safer and more effective drug co-prescription. Many computational approaches to predict the effect of DDIs have been proposed, with the aim of reducing the effort of identifying these interactions in vivo or in vitro, but room remains for improvement in prediction performance. RESULTS: In this study, we propose a novel deep learning model to predict the effect of DDIs more accurately.. The proposed model uses autoencoders and a deep feed-forward network that are trained using the structural similarity profiles (SSP), Gene Ontology (GO) term similarity profiles (GSP), and target gene similarity profiles (TSP) of known drug pairs to predict the pharmacological effects of DDIs. The results show that GSP and TSP increase the prediction accuracy when using SSP alone, and the autoencoder is more effective than PCA for reducing the dimensions of each profile. Our model showed better performance than the existing methods, and identified a number of novel DDIs that are supported by medical databases or existing research. CONCLUSIONS: We present a novel deep learning model for more accurate prediction of DDIs and their effects, which may assist in future research to discover novel DDIs and their pharmacological effects.


Assuntos
Aprendizado Profundo , Interações de Medicamentos , Modelos Teóricos , Área Sob a Curva , Bases de Dados Factuais , Humanos , Redes Neurais (Computação) , Máquina de Vetores de Suporte
9.
Stud Health Technol Inform ; 264: 1417-1418, 2019 Aug 21.
Artigo em Inglês | MEDLINE | ID: mdl-31438159

RESUMO

Automated wound detection has become a common issue in health care. A broad variety of image processing algorithms already exist, but they are very power consuming on mobile devices. Meanwhile the use of machine learning algorithms is on the rise and new frameworks have been developed to use these techniques with improved on-device-performance such as Apple Core Machine Learning Interface. In this paper, we evaluate the performance of libSVM for wound detection in practice.


Assuntos
Dermatopatias , Máquina de Vetores de Suporte , Algoritmos , Humanos , Processamento de Imagem Assistida por Computador , Aprendizado de Máquina
10.
Brain Nerve ; 71(7): 733-748, 2019 Jul.
Artigo em Japonês | MEDLINE | ID: mdl-31289247

RESUMO

Artificial intelligence (AI) shows promises in terms of diagnostic support on neuroimaging. We developed a software that predicts Alzheimer's disease (AD) using support vector machines (SVM) through three-dimensional brain MR images. Here, we will explain the general idea of voxel-based morphometry and SVM. We used the Alzheimer's Disease Neuroimaging Initiative (ADNI) database for SVM training and tested it on the Australian Imaging Biomarkers and Lifestyle Study of Ageing (AIBL) and the Japanese ADNI database. AI shows higher accuracy for predicting AD than a method of conventional statistical analysis, indicating potential clinical use for diagnostic support.


Assuntos
Doença de Alzheimer/diagnóstico por imagem , Encéfalo/diagnóstico por imagem , Neuroimagem , Máquina de Vetores de Suporte , Humanos
11.
Environ Monit Assess ; 191(8): 481, 2019 Jul 04.
Artigo em Inglês | MEDLINE | ID: mdl-31273539

RESUMO

This study presents a new fusion method namely supervised cross-fusion method to improve the capability of fused thermal, radar, and optical images for classification. The proposed cross-fusion method is a combination of pixel-based and supervised feature-based fusion of thermal, radar, and optical data. The pixel-based fusion was applied to fuse optical data of Sentinel-2 and Landsat 8. According to correlation coefficient (CR) and signal to noise ratio (SNR), among the used pixel-based fusion methods, wavelet obtained the best results for fusion. Considering spectral and spatial information preservation, CR of the wavelet method is 0.97 and 0.96, respectively. The supervised feature-based fusion method is a fusion of best output of pixel-based fusion level, land surface temperature (LST) data, and Sentinel-1 radar image using a supervised approach. The supervised approach is a supervised feature selection and learning of the inputs based on linear discriminant analysis and sparse regularization (LDASR) algorithm. In the present study, the non-negative matrix factorization (NMF) was utilized for feature extraction. A comparison of the obtained results with state of the art fusion method indicated a higher accuracy of our proposed method of classification. The rotation forest (RoF) classification results improvement was 25% and the support vector machine (SVM) results improvement was 31%. The results showed that the proposed method is well classified and separated four main classes of settlements, barren land, river, river bank, and even the bridges over the river. Also, a number of unclassified pixels by SVM are very low compared to other classification methods and can be neglected. The study results showed that LST calculated using thermal data has had positive effects on improving the classification results. By comparing the results of supervised cross-fusion without using LST data to the proposed method results, SVM and RoF classifiers showed 38% and 7% of classification improvement, respectively.


Assuntos
Monitoramento Ambiental/métodos , Processamento de Imagem Assistida por Computador/métodos , Algoritmos , Irã (Geográfico) , Radar , Rios , Máquina de Vetores de Suporte , Temperatura Ambiente
12.
J Food Sci ; 84(8): 2234-2241, 2019 Aug.
Artigo em Inglês | MEDLINE | ID: mdl-31313313

RESUMO

In order to rapidly and nondestructively identify tea grades, fluorescence hyperspectral imaging (FHSI) technology was proposed in this paper. A total of 309 Tieguanyin tea samples with three different grades were collected and the fluorescence hyperspectral data was acquired by hyperspectrometer (400 to 1000 nm). The characteristic wavelengths were respectively selected by Bootstrapping Soft Shrinkage (BOSS), Variable Iterative Space Shrinkage Approach (VISSA) and Model Adaptive Space Shrinkage (MASS) algorithms. Then, Support Vector Machine (SVM) was applied to establishing the relationship between the characteristic peaks, the full spectra, three characteristic spectra and the labels of tea grades. The results showed that VISSA-SVM model had the best classification performance, but the model precision can still be improved. Thus, Artificial Bee Colony (ABC) algorithm was introduced to optimize the parameters of SVM model. The accuracy and Kappa coefficient of test set of VISSA-ABC-SVM model were improved to 97.436% and 0.962, respectively. Therefore, the combination of fluorescence hyperspectra with VISSA-ABC-SVM model can accurately identify the grade of Tieguanyin tea. PRACTICAL APPLICATION: The rapid and accurate nondestructive tea grade identification method contributes to the construction of the tea online grade detection system. FHSI technology can solve the shortcomings of the reported methods and improved the identification accuracy of tea grades. It can be applied to the rapid detection of tea quality by tea companies, tea market, tea farmers and other demanders.


Assuntos
Algoritmos , Camellia sinensis/química , Imagem Óptica/métodos , Camellia sinensis/classificação , Análise Discriminante , Fluorescência , Folhas de Planta/química , Folhas de Planta/classificação , Máquina de Vetores de Suporte , Chá/química
13.
Stud Health Technol Inform ; 262: 344-347, 2019 Jul 04.
Artigo em Inglês | MEDLINE | ID: mdl-31349338

RESUMO

Clinical decision support systems are data analysis software that supports health professionals' decision - making the process to reach their ultimate outcome, taking into account patient information. However, the need for decision support systems cannot be denied because of most activities in the field of health care within the decision-making process. Decision support systems used for diagnosis are designed based on disease due to the complexity of diseases, symptoms, and disease-symptoms relationships. In the design and implementation of clinical decision support systems, mathematical modeling, pattern recognition and statistical analysis techniques of large databases and data mining techniques such as classification are also widely used. Classification of data is difficult in case of the small and/or imbalanced data set and this problem directly affects the classification performance. Small and/or imbalance dataset has become a major problem in data mining because classification algorithms are developed based on the assumption that the data sets are balanced and large enough. Most of the algorithms ignore or misclassify examples of the minority class, focus on the majority class. Most health data are small and imbalanced by nature. Learning from imbalanced and small data sets is an important and unsettled problem. Within the scope of the study, the publicly accessible data set, hepatitis was oversampled by distance-based data generation methods. The oversampled data sets were classified by using four different machine learning algorithms. Considering the classification scores of four different machine learning algorithms (Artificial Neural Networks, Support Vector Machines, Naive Bayes and Decision Tree), optimal synthetic data generation rate is recommended.


Assuntos
Algoritmos , Sistemas de Apoio a Decisões Clínicas , Aprendizado de Máquina , Teorema de Bayes , Humanos , Máquina de Vetores de Suporte
14.
Food Chem ; 295: 327-333, 2019 Oct 15.
Artigo em Inglês | MEDLINE | ID: mdl-31174765

RESUMO

Dual-pulse laser-induced breakdown spectroscopy (DPLIBS) and chemometric methods were used to predict chromium content in rice leaves, along with the purpose for increasing the detection sensitivity and accuracy. The influence of important parameters in DPLIBS were investigated and optimized. Then, partial least square (PLS) was used to establish chromium content prediction models, and the value of regression coefficient based on PLS was applied to determine feature variables. In addition, multivariate and univariate analysis were used to verify the modeling performance of selected feature variables. The results indicated that support vector machine model based on feature variables achieved the best performance, with correlation coefficient of 0.9946, root mean square error of 4.85 mg/kg and residual predictive deviation of 9.70 in prediction set. The proposed method provides a high-accuracy and fast approach for chromium content prediction in rice leaves, which could potentially be used for toxic and nutrient elements detection in food.


Assuntos
Cromo/análise , Oryza/química , Folhas de Planta/química , Análise Espectral/métodos , Poluentes Ambientais/análise , Lasers , Análise dos Mínimos Quadrados , Luz , Sensibilidade e Especificidade , Análise Espectral/estatística & dados numéricos , Máquina de Vetores de Suporte
15.
BMC Bioinformatics ; 20(Suppl 12): 314, 2019 Jun 20.
Artigo em Inglês | MEDLINE | ID: mdl-31216991

RESUMO

BACKGROUND: Microbiome profiles in the human body and environment niches have become publicly available due to recent advances in high-throughput sequencing technologies. Indeed, recent studies have already identified different microbiome profiles in healthy and sick individuals for a variety of diseases; this suggests that the microbiome profile can be used as a diagnostic tool in identifying the disease states of an individual. However, the high-dimensional nature of metagenomic data poses a significant challenge to existing machine learning models. Consequently, to enable personalized treatments, an efficient framework that can accurately and robustly differentiate between healthy and sick microbiome profiles is needed. RESULTS: In this paper, we propose MetaNN (i.e., classification of host phenotypes from Metagenomic data using Neural Networks), a neural network framework which utilizes a new data augmentation technique to mitigate the effects of data over-fitting. CONCLUSIONS: We show that MetaNN outperforms existing state-of-the-art models in terms of classification accuracy for both synthetic and real metagenomic data. These results pave the way towards developing personalized treatments for microbiome related diseases.


Assuntos
Algoritmos , Metagenômica/métodos , Redes Neurais (Computação) , Área Sob a Curva , Bases de Dados Genéticas , Humanos , Aprendizado de Máquina , Microbiota/genética , Modelos Teóricos , Fenótipo , Curva ROC , Máquina de Vetores de Suporte
16.
BMC Bioinformatics ; 20(1): 346, 2019 Jun 17.
Artigo em Inglês | MEDLINE | ID: mdl-31208321

RESUMO

BACKGROUND: Acetylation on lysine is a widespread post-translational modification which is reversible and plays a crucial role in some biological activities. To better understand the mechanism, it is necessary to identify acetylation sites in proteins accurately. Computational methods are popular because they are more convenient and faster than experimental methods. In this study, we proposed a new computational method to predict acetylation sites in human by combining sequence features and structural features including physicochemical property (PCP), position specific score matrix (PSSM), auto covariation (AC), residue composition (RC), secondary structure (SS) and accessible surface area (ASA), which can well characterize the information of acetylated lysine sites. Besides, a two-step feature selection was applied, which combined mRMR and IFS. It finally trained a cascade classifier based on SVM, which successfully solved the imbalance between positive samples and negative samples and covered all negative sample information. RESULTS: The performance of this method is measured with a specificity of 72.19% and a sensibility of 76.71% on independent dataset which shows that a cascade SVM classifier outperforms single SVM classifier. CONCLUSIONS: In addition to the analysis of experimental results, we also made a systematic and comprehensive analysis of the acetylation data.


Assuntos
Biologia Computacional/métodos , Máquina de Vetores de Suporte , Acetilação , Sequência de Aminoácidos , Animais , Bases de Dados de Proteínas , Ontologia Genética , Humanos , Lisina/química , Camundongos , Anotação de Sequência Molecular , Matrizes de Pontuação de Posição Específica , Processamento de Proteína Pós-Traducional , Estrutura Secundária de Proteína , Proteínas/química , Proteínas/metabolismo , Ratos
17.
Environ Monit Assess ; 191(7): 446, 2019 Jun 18.
Artigo em Inglês | MEDLINE | ID: mdl-31214787

RESUMO

Heavy metals in the agricultural soils of reclaimed mining areas can contaminate food and endanger human health. The objective of this study is to effectively estimate the concentrations of heavy metals, such as zinc, chromium, arsenic, and lead, using hyperspectral sensor data and the random forest (RF) algorithm in the study area of Xuzhou, China. The RF's built-in feature selection ability and modeling expressive ability in heavy metal estimation of soil were explored. After the preprocessing of the spectrum obtained by an ASD (analytical spectral device) field spectrometer, the random forest algorithm was carried out to establish the estimation model based on the correlation-selected features and the full-spectrum features respectively. Results of all the different processes were compared with classical approaches, such as partial least squares (PLS) regression and support vector machine (SVM). In all the experimental results, from the perspective of models, the best estimation model for Zn (R2 = 0.9061; RMSE = 6.5008) is based on the full-spectrum data of continuum removal (CR) pretreatment, and the best models for Cr (R2 = 0.9110; RMSE = 4.5683), As (R2 = 0.9912; RMSE = 0.5327), and Pb (R2 = 0.9756; RMSE = 1.1694) are all derived from the correlation-selected features. And these best models of these heavy metals are all established by the RF method. The experiments in this paper show that random forests can make full use of the input spectral data in the estimation of four kinds of heavy metals, and the obtained models are superior to those established by traditional methods.


Assuntos
Algoritmos , Arsênico/análise , Monitoramento Ambiental/estatística & dados numéricos , Metais Pesados/análise , Poluentes do Solo/análise , Agricultura , China , Monitoramento Ambiental/métodos , Humanos , Análise dos Mínimos Quadrados , Mineração , Máquina de Vetores de Suporte
18.
Food Chem ; 297: 124963, 2019 Nov 01.
Artigo em Inglês | MEDLINE | ID: mdl-31253305

RESUMO

Authentication of ground coffee has become an important issue because of fraudulent activities in the sector. In the current work, sixty-seven Brazilian coffees produced in different geographical origins using organic (ORG, n = 25) and conventional (CONV, n = 42) systems were analyzed for their stable isotope ratios (δ13C, δ18O, δ2H, and δ15N). Data were analyzed by inferential analysis to compare the factors whereas linear discriminant analysis (LDA), k-nearest neighbors (k-NN), and support vector machines (SVM) were used to classify the coffees based on their origin. ORG and CONV cultivated coffees could not be differentiated according to C stable isotope ratio (δ13C; p = 0.204), but ORG coffees presented higher values of the N stable isotope ratio (δ15N; p = 0.0006). k-NN presented the best classification results for both ORG and CONV coffees (87% and 67%, respectively). SVM correctly classified coffees produced in São Paulo (75% accuracy), while LDA correctly classified 71% of coffees produced in Minas Gerais.


Assuntos
Café/química , Análise de Alimentos/métodos , Espectrometria de Massas/métodos , Brasil , Isótopos de Carbono/análise , Deutério/análise , Análise Discriminante , Análise de Alimentos/estatística & dados numéricos , Espectrometria de Massas/estatística & dados numéricos , Isótopos de Nitrogênio/análise , Agricultura Orgânica , Isótopos de Oxigênio/análise , Máquina de Vetores de Suporte
19.
Accid Anal Prev ; 129: 202-210, 2019 Aug.
Artigo em Inglês | MEDLINE | ID: mdl-31170559

RESUMO

Detecting accidents is of great importance since they often impose significant delay and inconvenience to road users. This study compares the performance of two popular machine learning models, Support Vector Machine (SVM) and Probabilistic Neural Network (PNN), to detect the occurrence of accidents on the Eisenhower expressway in Chicago. Accordingly, since the detection of accidents should be as rapid as possible, seven models are trained and tested for each machine learning technique, using traffic condition data from 1 to 7 min after the actual occurrence. The main sources of data used in this study consist of weather condition, accident, and loop detector data. Furthermore, to overcome the problem of imbalanced data (i.e., underrepresentation of accidents in the dataset), the Synthetic Minority Oversampling TEchnique (SMOTE) is used. The results show that although SVM achieves overall higher accuracy, PNN outperforms SVM regarding the Detection Rate (DR) (i.e., percentage of correct accident detections). In addition, while both models perform best at 5 min after the occurrence of accidents, models trained at 3 or 4 min after the occurrence of an accident detect accidents more rapidly while performing reasonably well. Lastly, a sensitivity analysis of PNN for Time-To-Detection (TTD) reveals that the speed difference between upstream and downstream of accidents location is particularly significant to detect the occurrence of accidents.


Assuntos
Acidentes de Trânsito/estatística & dados numéricos , Redes Neurais (Computação) , Máquina de Vetores de Suporte , Chicago , Humanos , Fatores de Tempo , Tempo (Meteorologia)
20.
Food Chem ; 293: 213-219, 2019 Sep 30.
Artigo em Inglês | MEDLINE | ID: mdl-31151603

RESUMO

Soybean oil is often contaminated by aflatoxin B1 (AFB1) which is regarded as a class I carcinogen. The feasibility of rapid determination of AFB1 in soybean oil with terahertz spectroscopy was examined. t-SNE, as the pre-treatment method was used to get the best features and combined with different chemometrics including least squares-support vector machines (LS-SVM), back propagation neural network (BPNN), random forest (RF) and partial least squares (PLS) to find the best determination model. The excellent prediction results could be obtained using BPNN combined with t-SNE with correlation the coefficient of prediction (Rp) was 0.9948 and the root-mean-square error of prediction (RMSEP) was 0.7124 µg/kg. Besides, THz spectroscopy was proved to be feasible to detect AFB1 at 1 µg/kg in soybean oil (over 90% accuracy). It was concluded that THz spectroscopy together with chemometrics would be a promising technique for rapid determination of the AFB1 concentration in soybean oil.


Assuntos
Aflatoxina B1/análise , Óleo de Soja/química , Espectroscopia Terahertz/métodos , Análise dos Mínimos Quadrados , Análise de Componente Principal , Máquina de Vetores de Suporte
SELEÇÃO DE REFERÊNCIAS
DETALHE DA PESQUISA