RESUMO

Support Vector Machines (SVMs) are a type of supervised machine learning algorithm widely used for classification tasks. In contrast to traditional methods that split the data into separate training and testing sets, here we propose an innovative approach where subsets of the original data are randomly selected to train the model multiple times. This iterative training process aims to identify a representative data subset, leading to improved inferences about the population. Additionally, we introduce a novel distance-based kernel specifically designed for binary-type features based on a similarity matrix that efficiently handles both binary and multi-class classification problems. Computational experiments on publicly available datasets of varying sizes demonstrate that our proposed method significantly outperforms existing approaches in terms of classification accuracy. Furthermore, the distance-based kernel achieves superior performance compared to other well-known kernels from the literature and those used in previous studies on the same datasets. These findings validate the effectiveness of our proposed classification method and distance-based kernel for SVMs. By leveraging random subset selection and a unique kernel design, we achieve notable improvements in classification accuracy. These results have significant implications for diverse classification problems in Machine Learning and data analysis.

RESUMO

The development of methodologies to support decision-making in municipal solid waste (MSW) management processes is of great interest for municipal administrations. Artificial intelligence (AI) techniques provide multiple tools for designing algorithms to objectively analyze data while creating highly precise models. Support vector machines and neuronal networks are formed by AI applications offering optimization solutions at different managing stages. In this paper, an implementation and comparison of the results obtained by two AI methods on a solid waste management problem is shown. Support vector machine (SVM) and long short-term memory (LSTM) network techniques have been used. The implementation of LSTM took into account different configurations, temporal filtering and annual calculations of solid waste collection periods. Results show that the SVM method properly fits selected data and yields consistent regression curves, even with very limited training data, leading to more accurate results than those obtained by the LSTM method.

Assuntos

RESUMO

RESUMEN Introducción: La Anestesiología es la especialidad médica dedicada a la atención específica de los pacientes durante procedimientos quirúrgicos y en cuidados intensivos. Esta especialidad basada en los avances científicos y tecnológicos, ha incorporado el uso del monitoreo electroencefalográfico, facilitando el control continuo de estados de sedación anestésica durante las cirugías, con una adecuada concentración de fármacos. Objetivo: Proponer una estrategia de clasificación para el reconocimiento automático de tres estados de sedación anestésica en señales electroencefalográficas. Método: Se utilizaron con consentimiento informado escrito los registros electroencefalográficos de 27 pacientes sometidos a cirugía abdominal, excluyendo aquellos con antecedentes de epilepsia, enfermedades cerebrovasculares y otras afecciones neurológicas. Se aplicaron en total 12 fármacos anestésicos y dos relajantes musculares con montaje de 19 electrodos según el Sistema Internacional 10-20. Se eliminaron artefactos en los registros y se aplicaron técnicas de Inteligencia artificial para realizar el reconocimiento automático de los estados de sedación. Resultados: Se propuso una estrategia basada en el uso de máquinas de soporte vectorial con algoritmo multiclase Uno-Contra-Resto y la métrica Similitud Coseno, para realizar el reconocimiento automático de tres estados de sedación: profundo, moderado y ligero, en señales registradas por el canal frontal F4 y los occipitales O1 y O2. Se realizó una comparación de la propuesta con otros métodos de clasificación. Conclusiones: Se computa una exactitud balanceada del 92,67 % en el reconocimiento de los tres estados de sedación en las señales registradas por el canal electroencefalográfico F4, lo cual favorece el desarrollo de la monitorización anestésica.

ABSTRACT Introduction: Anesthesiology is the medical specialty concerned with the specific care of patients during surgical and intensive care procedures. This specialty, based on scientific and technological advances, has incorporated the use of electroencephalographic monitoring, facilitating the continuous control in the use of anesthesia for patient´s sedation states during surgeries, with an adequate concentration of drugs. Objective: Proposal for a classification strategy for automatic recognition of three sedation states in electroencephalographic signals. Methods: We used, with written informed consent, the electroencephalographic records of 27 patients undergoing abdominal surgery, excluding those with a history of epilepsy, cerebrovascular disease and other neurological conditions. A total of 12 drugs to produce anesthesia and two muscle relaxants with 19 electrodes, mounted according to the International System 10 -20, were applied. Artifacts in the records were eliminated and artificial intelligence techniques were applied to perform automatic recognition of sedation states. Results: A strategy based on the use of support vector machines with a multiclass algorithm One-against-Rest and the Cosine Similarity metric was proposed to perform the automatic recognition of three sedation states: deep, moderate and light, in signals recorded by the frontal channel F4 and the occipital channels O1 and O2. A comparison was carried out between the proposal showed and other classification methods. Conclusions: A balanced accuracy of 92.67% is computed about the recognition of the three states of sedation in the signals recorded by the electroencephalographic channel F4, which helps in a better anesthetic monitoring process.

RESUMO Introdução: A Anestesiologia é a especialidade médica dedicada ao atendimento específico de pacientes durante procedimentos cirúrgicos e em terapia intensiva. Essa especialidade, baseada nos avanços científicos e tecnológicos, incorporou o uso da monitorização eletroencefalográfica, facilitando o controle contínuo dos estados de sedação anestésica durante as cirurgias, com concentração adequada de fármacos. Objetivo: Propor uma estratégia de classificação para o reconhecimento automático de três estados de sedação anestésica em sinais eletroencefalográficos. Método: Foram utilizados registros eletroencefalográficos de 27 pacientes submetidos à cirurgia abdominal com consentimento informado por escrito, excluindo aqueles com histórico de epilepsia, doenças cerebrovasculares e outras condições neurológicas. Um total de 12 drogas anestésicas e dois relaxantes musculares foram aplicados com um conjunto de 19 eletrodos de acordo com o Sistema Internacional 10-20. Artefatos nos prontuários foram removidos e técnicas de inteligência artificial foram aplicadas para realizar o reconhecimento automático dos estados de sedação. Resultados: Foi proposta uma estratégia baseada no uso de máquinas de vetores de suporte com algoritmo One-Against-Rest multiclasse e a métrica Cosine Similarity para realizar o reconhecimento automático de três estados de sedação: profundo, moderado e leve, em sinais registrados pelo canal frontal F4 e os canais occipitais O1 e O2. Foi feita uma comparação da proposta com outros métodos de classificação. Conclusões: Uma acurácia equilibrada de 92,67% é computada no reconhecimento dos três estados de sedação nos sinais registrados pelo canal eletroencefalográfico F4, o que favorece o desenvolvimento da monitorização anestésica.

RESUMO

This paper presents new approaches to fit regression models for symbolic internal-valued variables, which are shown to improve and extend the center method suggested by Billard and Diday and the center and range method proposed by Lima-Neto, E.A.and De Carvalho, F.A.T. Like the previously mentioned methods, the proposed regression models consider the midpoints and half of the length of the intervals as additional variables. We considered various methods to fit the regression models, including tree-based models, K-nearest neighbors, support vector machines, and neural networks. The approaches proposed in this paper were applied to a real dataset and to synthetic datasets generated with linear and nonlinear relations. For an evaluation of the methods, the root-mean-squared error and the correlation coefficient were used. The methods presented herein are available in the the RSDA package written in the R language, which can be installed from CRAN.

RESUMO

Using near-infrared (NIR) spectroscopy for poultry litter characterization can be a rapid, non-destructive, and low-cost alternative. This study aims to estimate the C, N, P, and K content in poultry litter samples using for first time NIR spectroscopy. For these purposes, the building models were carried out using Partial Least Squares (PLS) and Support Vector Machines (SVM) methods. A total of 160 litter samples were analyzed in poultry houses of different rearing systems, seeking the highest possible variability in their chemical composition. NIR spectroscopy, combined with PLS and SVM methods, is an alternative method for non-destructive C, N, P, and K determination in poultry samples. The regression models using SVM provide better accuracy for all elements, laying the basis for the nonlinear regression approach's application. The K determination on poultry litter using NIR was possible only by the SVM model (R2 = 0.8620 and RPD = 2.7330). Conclusively, the predictive ability was improved using the SVM method.

RESUMO

BACKGROUND AND OBJECTIVE: Pressure ulcers are regions of trauma caused by a continuous pressure applied to soft tissues between a bony prominence and a hard surface. The manual monitoring of their healing evolution can be achieved by area assessment techniques that include the use of rulers and adhesive labels in direct contact with the injury, being highly inaccurate and subjective. In this paper we present a Support Vector Machine classifier in combination with a modified version of the GrabCut method for the automatic measurement of the area affected by pressure ulcers in digital images. METHODS: Three methods of region segmentation using the superpixel strategy were evaluated from which color and texture descriptors were extracted. After the superpixel classification, the GrabCut segmentation method was applied in order to delineate the region affected by the ulcer from the rest of the image. RESULTS: Experiments on a set of 105 pressure ulcer images from a public data set resulted in an average accuracy of 96%, sensitivity of 94%, specificity of 97% and precision of 94%. CONCLUSIONS: The association of support vector machines with superpixel segmentation outperformed current methods based on deep learning and may be extended to tissue classification.

Assuntos

RESUMO

A discovery-based lipid profiling study of serum samples from a cohort that included patients with clear cell renal cell carcinoma (ccRCC) stages I, II, III, and IV (n = 112) and controls (n = 52) was performed using ultraperformance liquid chromatography coupled to quadrupole-time-of-flight mass spectrometry and machine learning techniques. Multivariate models based on support vector machines and the LASSO variable selection method yielded two discriminant lipid panels for ccRCC detection and early diagnosis. A 16-lipid panel allowed discriminating ccRCC patients from controls with 95.7% accuracy in a training set under cross-validation and 77.1% accuracy in an independent test set. A second model trained to discriminate early (I and II) from late (III and IV) stage ccRCC yielded a panel of 26 compounds that classified stage I patients from an independent test set with 82.1% accuracy. Thirteen species, including cholic acid, undecylenic acid, lauric acid, LPC(16:0/0:0), and PC(18:2/18:2), identified with level 1 exhibited significantly lower levels in samples from ccRCC patients compared to controls. Moreover, 3α-hydroxy-5α-androstan-17-one 3-sulfate, cis-5-dodecenoic acid, arachidonic acid, cis-13-docosenoic acid, PI(16:0/18:1), PC(16:0/18:2), and PC(O-16:0/20:4) contributed to discriminate early from late ccRCC stage patients. The results are auspicious for early ccRCC diagnosis after validation of the panels in larger and different cohorts.

Assuntos

RESUMO

The correct classification of requirements has become an essential task within software engineering. This study shows a comparison among the text feature extraction techniques, and machine learning algorithms to the problem of requirements engineer classification to answer the two major questions "Which works best (Bag of Words (BoW) vs. Term Frequency-Inverse Document Frequency (TF-IDF) vs. Chi Squared (CHI2)) for classifying Software Requirements into Functional Requirements (FR) and Non-Functional Requirements (NF), and the sub-classes of Non-Functional Requirements?" and "Which Machine Learning Algorithm provides the best performance for the requirements classification task?". The data used to perform the research was the PROMISE_exp, a recently made dataset that expands the already known PROMISE repository, a repository that contains labeled software requirements. All the documents from the database were cleaned with a set of normalization steps and the two feature extractions, and feature selection techniques used were BoW, TF-IDF and CHI2 respectively. The algorithms used for classification were Logist Regression (LR), Support Vector Machine (SVM), Multinomial Naive Bayes (MNB) and k-Nearest Neighbors (kNN). The novelty of our work is the data used to perform the experiment, the details of the steps used to reproduce the classification, and the comparison between BoW, TF-IDF and CHI2 for this repository not having been covered by other studies. This work will serve as a reference for the software engineering community and will help other researchers to understand the requirement classification process. We noticed that the use of TF-IDF followed by the use of LR had a better classification result to differentiate requirements, with an F-measure of 0.91 in binary classification (tying with SVM in that case), 0.74 in NF classification and 0.78 in general classification. As future work we intend to compare more algorithms and new forms to improve the precision of our models.

RESUMO

Condensed matter physics (CMP) seeks to understand the microscopic interactions of matter at the quantum and atomistic levels, and describes how these interactions result in both mesoscopic and macroscopic properties. CMP overlaps with many other important branches of science, such as chemistry, materials science, statistical physics, and high-performance computing. With the advancements in modern machine learning (ML) technology, a keen interest in applying these algorithms to further CMP research has created a compelling new area of research at the intersection of both fields. In this review, we aim to explore the main areas within CMP, which have successfully applied ML techniques to further research, such as the description and use of ML schemes for potential energy surfaces, the characterization of topological phases of matter in lattice systems, the prediction of phase transitions in off-lattice and atomistic simulations, the interpretation of ML theories with physics-inspired frameworks and the enhancement of simulation methods with ML algorithms. We also discuss in detail the main challenges and drawbacks of using ML methods on CMP problems, as well as some perspectives for future developments.

RESUMO

Advancement on computer and sensing technologies has generated exponential growth in the data available for the development of systems that support decision-making in fields such as health, entertainment, manufacturing, among others. This fact has made that the fusion of data from multiple and heterogeneous sources became one of the most promising research fields in machine learning. However, in real-world applications, to reduce the number of sources while maintaining optimal system performance is an important task due to the availability of data and implementation costs related to processing, implementation, and development times. In this work, a novel method for the objective selection of relevant information sources in a multimodality system is proposed. This approach takes advantage of the ability of multiple kernel learning (MKL) and the support vector machines (SVM) classifier to perform an optimal fusion of data by assigning weights according to their discriminative value in the classification task; when a kernel is designed for representing each data source, these weights can be used as a measure of their relevance. Moreover, three algorithms for tuning the Gaussian kernel bandwidth in the classifier prediction stage are introduced to reduce the computational cost of searching for an optimal solution; these algorithms are an adaptation of a common technique in unsupervised learning named local scaling. Two real application tasks were used to evaluate the proposed method: the selection of electrodes for a classification task in Brain-Computer Interface (BCI) systems and the selection of relevant Magnetic Resonance Imaging (MRI) sequences for detection of breast cancer. The obtained results show that the proposed method allows the selection of a small number of information sources.

RESUMO

Previous real-time crash prediction models have scarcely used data disaggregated by vehicle type such as light, heavy and motorcycles. Thus, little effort has been made to quantify the impact of flow composition variables as crash precursors. We analyze the advantages of having access to this data by analyzing two scenarios, namely, with aggregated and disaggregated data. For each case, we build Logistics Regressions and Support Vector Machines models to predict accidents in a major urban expressway in Santiago, Chile. Our results show that having access to disaggregated data by vehicle type increases the prediction power up to 30 % providing, at the same time, much better intuition about the actual traffic conditions that may lead to accidents. These results may be useful when evaluating technology investments and developments in urban freeways.

Assuntos

RESUMO

This study presents an analysis of three models associated with artificial intelligence as tools to forecast the generation of urban solid waste in the city of Bogotá, in order to learn about this type of waste's behavior. The analysis was carried out in such a manner that different efficient alternatives are presented. In this paper, a possible decision-making strategy was explored and implemented to plan and design technologies for the stages of collection, transport and final disposal of waste in cities, while taking into account their particular characteristics. The first model used to analyze data was the decision tree which employed machine learning as a non-parametric algorithm that models data separation limitations based on the learning decision rules on the input characteristics of the model. Support vector machines were the second method implemented as a forecasting model. The primary advantage of support vector machines is their proper adjustment to data despite its variable nature or when faced with problems with a small amount of training data. Lastly, recurrent neural network models to forecast data were implemented, which yielded positive results. Their architectural design is useful in exploring temporal correlations among the same. Distribution by collection zone in the city, socio-economic stratification, population, and quantity of solid waste generated in a determined period of time were factors considered in the analysis of this forecast. The results found that support vector machines are the most appropriate model for this type of analysis.

RESUMO

The demand for information on the soil resource to support the establishment of public policies for land use and management has grown exponentially in the last years. However, there are still difficulties to the proper use of already existing information for soil mapping. Here we aimed to establish a protocol for soil mapping using legacy data, magnetic signature and soil attributes evaluation. A total of 493 soil samples were collected at 0-0.20 m in the geological domain of Western Plateau of São Paulo State. This work has three parts: First, we performed a classification analysis using soil mapping units (SMU) extracted from conventional soil map and Support Vector Machines algorithm (SVM). As covariates, we used categorical information, such as geology, dissection and landform maps. Second, we used soil attributes to perform a cluster analysis using k-means as partitioning method. To choose the optimal number of clusters, the same number of SMU showed in the conventional soil map (e.g. 34 clusters) were used. The last step was to compare soil and clusters maps predicted by SVM with the conventional soil map. Results showed good performance of SVM for both classifications (clusters and SMU), with overall accuracy of 0.60 and 0.90 respectively. In addition, the distribution of soil attributes within each cluster was more homogeneous and well distributed than within SMU, showing that is very possible to use numerical classification for soil mapping. Future soil surveys could use cluster analysis as a preliminary evaluation for better understanding of tropical soil variations.

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

RESUMO

INTRODUCTION: Novel drug discovery remains an enormous challenge, with various computer-aided drug design (CADD) approaches having been widely employed for this purpose. CADD, specifically the commonly used support vector machines (SVMs), can employ machine learning techniques. SVMs and their variations offer numerous drug discovery applications, which range from the classification of substances (as active or inactive) to the construction of regression models and the ranking/virtual screening of databased compounds. Areas covered: Herein, the authors consider some of the applications of SVMs in medicinal chemistry, illustrating their main advantages and disadvantages, as well as trends in their utilization, via the available published literature. The aim of this review is to provide an up-to-date review of the recent applications of SVMs in drug discovery as described by the literature, thereby highlighting their strengths, weaknesses, and future challenges. Expert opinion: Techniques based on SVMs are considered as powerful approaches in early drug discovery. The ability of SVMs to classify active or inactive compounds has enabled the prioritization of substances for virtual screening. Indeed, one of the main advantages of SVMs is related to their potential in the analysis of nonlinear problems. However, despite successes in employing SVMs, the challenges of improving accuracy remain.

Assuntos

RESUMO

Computational self-adapting methods (Support Vector Machines, SVM) are compared with an analytical method in effluent composition prediction of a two-stage anaerobic digestion (AD) process. Experimental data for the AD of poultry manure were used. The analytical method considers the protein as the only source of ammonia production in AD after degradation. Total ammonia nitrogen (TAN), total solids (TS), chemical oxygen demand (COD), and total volatile solids (TVS) were measured in the influent and effluent of the process. The TAN concentration in the effluent was predicted, this being the most inhibiting and polluting compound in AD. Despite the limited data available, the SVM-based model outperformed the analytical method for the TAN prediction, achieving a relative average error of 15.2% against 43% for the analytical method. Moreover, SVM showed higher prediction accuracy in comparison with Artificial Neural Networks. This result reveals the future promise of SVM for prediction in non-linear and dynamic AD processes. Graphical abstract ᅟ.

Assuntos

RESUMO

This article describes a methodology to recognize emotional states through an electroencephalography signals analysis, developed with the premise of reducing the computational burden that is associated with it, implementing a strategy that reduces the amount of data that must be processed by establishing a relationship between electrodes and Brodmann regions, so as to discard electrodes that do not provide relevant information to the identification process. Also some design suggestions to carry out a pattern recognition process by low computational complexity neural networks and support vector machines are presented, which obtain up to a 90.2% mean recognition rate.

Assuntos

RESUMO

Epilepsy is a neurological disorder that causes changes in the autonomic nervous system. Heart rate variability (HRV) reflects the regulation of cardiac activity and autonomic nervous system tone. The early detection of epileptic seizures could foster the use of new treatment approaches. This study presents a new methodology for the prediction of epileptic seizures using HRV signals. Eigendecomposition of HRV parameter covariance matrices was used to create an input for a support vector machine (SVM)-based classifier. We analyzed clinical data from 12 patients (9 female; 3 male; age 34.5 ± 7.5 years), involving 34 seizures and a total of 55.2 h of interictal electrocardiogram (ECG) recordings. Data from 123.6 h of ECG recordings from healthy subjects were used to test false positive rate per hour (FP/h) in a completely independent data set. Our methodological approach allowed the detection of impending seizures from 5 min to just before the onset of a clinical/electrical seizure with a sensitivity of 94.1%. The FP rate was 0.49 h-1 in the recordings from patients with epilepsy and 0.19 h-1 in the recordings from healthy subjects. Our results suggest that it is feasible to use the dynamics of HRV parameters for the early detection and, potentially, the prediction of epileptic seizures.

RESUMO

BACKGROUND: Geminiviruses infect a broad range of cultivated and non-cultivated plants, causing significant economic losses worldwide. The studies of the diversity of species, taxonomy, mechanisms of evolution, geographic distribution, and mechanisms of interaction of these pathogens with the host have greatly increased in recent years. Furthermore, the use of rolling circle amplification (RCA) and advanced metagenomics approaches have enabled the elucidation of viromes and the identification of many viral agents in a large number of plant species. As a result, determining the nomenclature and taxonomically classifying geminiviruses turned into complex tasks. In addition, the gene responsible for viral replication (particularly, the viruses belonging to the genus Mastrevirus) may be spliced due to the use of the transcriptional/splicing machinery in the host cells. However, the current tools have limitations concerning the identification of introns. RESULTS: This study proposes a new method, designated Fangorn Forest (F2), based on machine learning approaches to classify genera using an ab initio approach, i.e., using only the genomic sequence, as well as to predict and classify genes in the family Geminiviridae. In this investigation, nine genera of the family Geminiviridae and their related satellite DNAs were selected. We obtained two training sets, one for genus classification, containing attributes extracted from the complete genome of geminiviruses, while the other was made up to classify geminivirus genes, containing attributes extracted from ORFs taken from the complete genomes cited above. Three ML algorithms were applied on those datasets to build the predictive models: support vector machines, using the sequential minimal optimization training approach, random forest (RF), and multilayer perceptron. RF demonstrated a very high predictive power, achieving 0.966, 0.964, and 0.995 of precision, recall, and area under the curve (AUC), respectively, for genus classification. For gene classification, RF could reach 0.983, 0.983, and 0.998 of precision, recall, and AUC, respectively. CONCLUSIONS: Therefore, Fangorn Forest is proven to be an efficient method for classifying genera of the family Geminiviridae with high precision and effective gene prediction and classification. The method is freely accessible at www.geminivirus.org:8080/geminivirusdw/discoveryGeminivirus.jsp .

Assuntos

RESUMO

For all species, abiotic factors directly affect performance, survival and reproduction, and consequently, their geographic distribution. Species distribution models (SDMs) are important tools to predict the influence of abiotic factors in species distributions and has been more applied over the years. However, these models can be built under different algorithms and using different methods to select environmental predictors, which can lead to different results. Five different algorithms and two sets of environmental predictors were compared to predict the geographic distribution of the blowfly Sarconesia chlorogaster (Wiedemann) (Diptera: Calliphoridae). This species has several occurrence points and a considerable amount of biological data available, which makes S. chlorogaster a good model system to compare environmental predictors. Two sets of environmental predictors (mainly derived from temperature and humidity) were built, and the set based on the influence of abiotic variables on the ecophysiology of S. chlorogaster showed better results than the principal component analysis (PCA) approach using 19 climatic variables. We also employed five modeling algorithms-Envelope Score, Mahalanobis Distance, GARP, Support Vector Machines, and Maxent-and the latter two showed the best performances. The results indicate that temperature is the main factor shaping geographic distribution of S. chlorogaster through its effect on fitness. Furthermore, we showed that this species is mainly distributed in south, southeastern, and some northwestern and southwestern sites of South America. In addition, our results also predicted suitable areas in Ecuador and Colombia, countries without previous records.

Assuntos