Academic literature on the topic 'GMM-EM'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the lists of relevant articles, books, theses, conference reports, and other scholarly sources on the topic 'GMM-EM.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Journal articles on the topic "GMM-EM"
1
Gu, Ming Liang, and Biao Zhang. "Chinese Dialect Identification Using SC-GMM." Advanced Materials Research 433-440 (January 2012): 3292–96. http://dx.doi.org/10.4028/www.scientific.net/amr.433-440.3292.
Full textAbstract:
Gaussian mixture model (GMM) is a sort of effective identification method in Chinese dialects identification, estimating GMM parameters is always an important step in building a state-of-the-art speech processing system. One of the most widely used approaches is maximum- likelihood estimation, where parameters of class-specific distributions are estimated using Expectation Maximization algorithm(EM). Initial parameters have great influence on the convergence of EM algorithm, so how to initialize GMM parameters is a key problem. In this paper, we apply spectral clustering(SC) to initialize GMM parameters. Experimental results prove that using spectral clustering algorithm to initialize GMM parameters is superior to traditional K-Means method and identification system has a higher recognition rate.
2
Li, Yupeng, Jianhua Zhang, Ruisi He, Lei Tian, and Hewen Wei. "Hybrid DE-EM Algorithm for Gaussian Mixture Model-Based Wireless Channel Multipath Clustering." International Journal of Antennas and Propagation 2019 (December 27, 2019): 1–10. http://dx.doi.org/10.1155/2019/4639612.
Full textAbstract:
In this paper, the Gaussian mixture model (GMM) is introduced to the channel multipath clustering. In the GMM field, the expectation-maximization (EM) algorithm is usually utilized to estimate the model parameters. However, the EM widely converges into local optimization. To address this issue, a hybrid differential evolution (DE) and EM (DE-EM) algorithms are proposed in this paper. To be specific, the DE is employed to initialize the GMM parameters. Then, the parameters are estimated with the EM algorithm. Thanks to the global searching ability of DE, the proposed hybrid DE-EM algorithm is more likely to obtain the global optimization. Simulations demonstrate that our proposed DE-EM clustering algorithm can significantly improve the clustering performance.
3
Xu, Yanping, Tingcong Ye, Xin Wang, Yuping Lai, Jian Qiu, Lingjun Zhang, and Xia Zhang. "GMM with parameters initialization based on SVD for network threat detection." Journal of Intelligent & Fuzzy Systems 40, no. 1 (January 4, 2021): 477–90. http://dx.doi.org/10.3233/jifs-200066.
Full textAbstract:
In the field of security, the data labels are unknown or the labels are too expensive to label, so that clustering methods are used to detect the threat behavior contained in the big data. The most widely used probabilistic clustering model is Gaussian Mixture Models(GMM), which is flexible and powerful to apply prior knowledge for modelling the uncertainty of the data. Therefore, in this paper, we use GMM to build the threat behavior detection model. Commonly, Expectation Maximization (EM) and Variational Inference (VI) are used to estimate the optimal parameters of GMM. However, both EM and VI are quite sensitive to the initial values of the parameters. Therefore, we propose to use Singular Value Decomposition (SVD) to initialize the parameters. Firstly, SVD is used to factorize the data set matrix to get the singular value matrix and singular matrices. Then we calculate the number of the components of GMM by the first two singular values in the singular value matrix and the dimension of the data. Next, other parameters of GMM, such as the mixing coefficients, the mean and the covariance, are calculated based on the number of the components. After that, the initialization values of the parameters are input into EM and VI to estimate the optimal parameters of GMM. The experiment results indicate that our proposed method performs well on the parameters initialization of GMM clustering using EM and VI for estimating parameters.
4
Tilley, Roger, Hamid Sadjadpour, and Farid Dowla. "GPR Imaging for Deeply Buried Objects: A Comparative Study Based on Compositing of Scanning Frequencies and a Chirp Excitation Function." Geosciences 9, no. 3 (March 18, 2019): 132. http://dx.doi.org/10.3390/geosciences9030132.
Full textAbstract:
Compositing of ground penetrating radar (GPR) scans of differing frequencies have been found to produce cleaner images at depth using the Gaussian mixture model (GMM) feature of the expectation-maximization (EM) algorithm. GPR scans at various heights (“Stand Off”), as well as ground-based scans, have been studied. In this paper, we compare the GPR response from a chirp excitation function-based radar with the response from the EM GMM algorithm compositing process, using the same mix of frequencies. A chirp excitation pulse was found to be effective in delineating the defined buried object, but the resulting image is less sharp than the GMM EM method.
5
Pudurkar, Rahul, Shruti Patil, Gazala Ansari, and Shaikh Phiroj. "Biometric Voice Recognition system using MFCC and GMM with EM." International Journal of Computer Applications 184, no. 26 (August 20, 2022): 5–10. http://dx.doi.org/10.5120/ijca2022922315.
Full text
6
Ma, Bo, Mingyang Wu, Zhilu Wu, Zhendong Yin, and Tao Shen. "GMM-MUD: An Effective Multiuser Detection Algorithm for DS-UWB-Based Space Formation Flying Systems." Mathematical Problems in Engineering 2019 (November 20, 2019): 1–10. http://dx.doi.org/10.1155/2019/4350794.
Full textAbstract:
In this paper, an effective multiuser detection (MUD) is proposed for direct sequence ultrawideband- (DS-UWB-) based space formation flying systems. The proposed method called GMM-MUD is based on Gaussian mixture models (GMMs) to suppress multiple access interference. The GMM describes probability distributions of the hypothesis testing problem which is used for bit classification. To reveal the difference between correct bits and error bits, the preprocessing operation applies a mapping function based on optimal multiuser detection. The parameters of GMM are estimated by using expectation-maximization (EM) algorithm. EM algorithm employs iterative operation to simplify the complexity of maximum likelihood estimation method and considers the mapping values of received bits as the observations. Simulation results demonstrate that the proposed GMM-MUD algorithm achieves good performances in terms of bit error rate performance, user capacity, and near-far resistance. Moreover, the computational complexity is low enough for space formation flying applications.
7
Hashimoto, Juliano Hideo, Claudete Regina Alcalde, Maximiliane Alavarse Zambom, Karina Toledo da Silva, Francisco de Assis Fonseca de Macedo, Elias Nunes Martins, Carlos Eduardo Crispim de Oliveira Ramos, and Gabriella de Oliveira Passianoto. "Desempenho e digestibilidade aparente em cabritos Boer x Saanen em confinamento recebendo rações com casca do grão de soja em substituição ao milho." Revista Brasileira de Zootecnia 36, no. 1 (February 2007): 174–82. http://dx.doi.org/10.1590/s1516-35982007000100021.
Full textAbstract:
Foram utilizados 27 cabritos Boer x Saanen (15 machos e 12 fêmeas, 25,75 ± 3,80 kg PV) distribuídos em delineamento inteiramente casualizado para avaliar os os efeitos da substituição de casca do grão de soja (CGS) por grão de milho moído (GMM) sobre o desempenho e a digestibilidade aparente dos nutrientes da dieta. As rações foram compostas de feno de grama-estrela (Cynodon spp.), farelo de soja, milho, casca do grão de soja e minerais, sendo os tratamentos constituídos de 0 (GMM), 50 (CGS50) e 100% (CGS100) de substituição do milho pela casca do grão de soja. O período experimental foi de 42 dias e a digestibilidade dos nutrientes foi determinada utilizando-se a FDN indigestível como indicador. Os tratamentos não diferiram quanto às ingestões de MS, MO e PB, ao ganho de peso médio diário e à conversão alimentar. A utilização da CGS resultou em diminuição dos coeficientes de digestibilidade da MS, MO e PB e aumento dos coeficientes de digestibilidade da FDN e do amido. A digestibilidade in vitro da MS foi maior no tratamento GMM, mas a da parede celular não diferiu entre os tratamentos. A casca do grão de soja apresentou 23,87; 0,19; 55,89 e 20,05% das frações A, B1, B2 e C dos carboidratos e 35,22; 8,19; 31,61; 15,46 e 9,52% das frações A, B1, B2, B3 e C da proteína. A casca do grão de soja, quando utilizada em até 100% de substituição ao milho nas rações, não altera o desempenho de cabritos em crescimento, mas diminui a digestibilidade da MS, MO e da proteína e aumenta a da FDN.
8
Sun, Qi, Liwen Jiang, and Haitao Xu. "Expectation-Maximization Algorithm of Gaussian Mixture Model for Vehicle-Commodity Matching in Logistics Supply Chain." Complexity 2021 (January 13, 2021): 1–11. http://dx.doi.org/10.1155/2021/9305890.
Full textAbstract:
A vehicle-commodity matching problem (VCMP) is presented for service providers to reduce the cost of the logistics system. The vehicle classification model is built as a Gaussian mixture model (GMM), and the expectation-maximization (EM) algorithm is designed to solve the parameter estimation of GMM. A nonlinear mixed-integer programming model is constructed to minimize the total cost of VCMP. The matching process between vehicle and commodity is realized by GMM-EM, as a preprocessing of the solution. The design of the vehicle-commodity matching platform for VCMP is designed to reduce and eliminate the information asymmetry between supply and demand so that the order allocation can work at the right time and the right place and use the optimal solution of vehicle-commodity matching. Furthermore, the numerical experiment of an e-commerce supply chain proves that a hybrid evolutionary algorithm (HEA) is superior to the traditional method, which provides a decision-making reference for e-commerce VCMP.
9
Pelizza, Cristian Rafael. "UMA ANÁLISE DO CANAL DE TRANSMISSÃO E DA DEMANDA AGREGADA POR CRÉDITO NO BRASIL, DE 2003 A 2012." Revista Cadernos de Economia 21, no. 38 (March 12, 2019): 63–73. http://dx.doi.org/10.46699/rce.v21i38.4299.
Full textAbstract:
O objetivo do presente artigo é apresentar um modelo, com base no trabalho de Bernanke e Blinder (1988), em que o crédito possui impacto macroeconômico, como canal de transmissão das políticas monetárias, bem como estimar a demanda agregada por crédito, observando o impacto da taxa Selic sobre a mesma. O ajuste do modelo econométrico foi feito utilizando os estimadores de Mínimos Quadrados em Dois Estágios (MQ2E) e Método Generalizado dos Momentos (GMM), sendo as variáveis instrumentais utilizadas relacionadas à oferta agregada de crédito. Os resultados são congruentes com a teoria econômica, com o crescimento de crédito sendo de -0,35% (MQ2E) e -0,55% (GMM) para uma variação de 1% na Selic.
10
OSOBA, OSONDE, SANYA MITAIM, and BART KOSKO. "THE NOISY EXPECTATION–MAXIMIZATION ALGORITHM." Fluctuation and Noise Letters 12, no. 03 (September 2013): 1350012. http://dx.doi.org/10.1142/s0219477513500120.
Full textAbstract:
We present a noise-injected version of the expectation–maximization (EM) algorithm: the noisy expectation–maximization (NEM) algorithm. The NEM algorithm uses noise to speed up the convergence of the EM algorithm. The NEM theorem shows that additive noise speeds up the average convergence of the EM algorithm to a local maximum of the likelihood surface if a positivity condition holds. Corollary results give special cases when noise improves the EM algorithm. We demonstrate these noise benefits on EM algorithms for three data models: the Gaussian mixture model (GMM), the Cauchy mixture model (CMM), and the censored log-convex gamma model. The NEM positivity condition simplifies to a quadratic inequality in the GMM and CMM cases. A final theorem shows that the noise benefit for independent identically distributed additive noise decreases with sample size in mixture models. This theorem implies that the noise benefit is most pronounced if the data is sparse.
Dissertations / Theses on the topic "GMM-EM"
1
Sasseron, Ricardo Henrique. "Determinantes da produtividade: análise do impacto do índice GCI e seus componentes sobre a PTF." Universidade de São Paulo, 2016. http://www.teses.usp.br/teses/disponiveis/96/96131/tde-20102016-140527/.
Full textAbstract:
Este trabalho buscou investigar os principais determinantes da produtividade. Para tanto, o trabalho construiu medidas da Produtividade Total dos Fatores (PTF) e analisou o impacto do Global Competitiveness Index (GCI), de seus subíndices e dos 12 pilares de competitividade que compõem esses subíndices, sobre a PTF. Os principais resultados encontrados são: a heterogeneidade não observada dos países é correlacionada com os regressores, causando estimativas inconsistentes caso não seja tratada; as variáveis apresentaram endogeneidade nos modelos, a qual também deve ser levada em consideração nas estimações; a produtividade passada se mostrou relevante para explicar seu comportamento futuro; os pilares \"ambiente macroeconômico\", \"saúde e ensino básico\" e \"desenvolvimento do mercado financeiro\" afetam positivamente a PTF; já a \"capacidade de absorção tecnológica\" apresentou efeito líquido negativo sobre a PTF, ao contrário do que era esperado; os subíndices apresentaram poucos resultados estatisticamente significantes e em linha com o esperado; e, por último, a análise do índice GCI revelou que um aumento em uma unidade desse indicador, em uma escala de 1 a 7, está relacionado a um crescimento líquido de 21% na produtividadeThis study aimed to investigate the main determinants of productivity. Therefore, the work built measures of Total Factor Productivity (TFP) and analyzed the impact of the Global Competitiveness Index (GCI), its subindexes and the 12 pillars of competitiveness that make up these subindexes on the PTF. The main results are: the unobserved heterogeneity of countries is related to the covariates, causing inconsistent estimates if not treated; the variables presented endogeneity in the models, which should also be taken into account in the estimation; productivity lags proved relevant to explain their future behavior; the pillars \"macroeconomic environment\", \"health and basic education\" and \"financial market development\" positively affect the TFP; on the other hand, \"technological readiness\" had a negative net effect on TFP, contrary to what was expected; the subindexes showed few statistically significant results and in line with what was expected; and finally, the analysis of the GCI index revealed that an increase in one unit of this indicator, on a 1 to 7 scale, is related to a net increase of 21% in productivity
2
Goulart, Antonio José Homsi. "Classificação automática de gênero musical baseada em entropia e fractais." Universidade de São Paulo, 2012. http://www.teses.usp.br/teses/disponiveis/18/18153/tde-16032012-152801/.
Full textAbstract:
A classificação automática de gênero musical tem como finalidade o conforto de ouvintes de músicas auxiliando no gerenciamento das coleções de músicas digitais. Existem sistemas que se baseiam em cabeçalhos de metadados (tais como nome de artista, gênero cadastrado, etc.) e também os que extraem parâmetros dos arquivos de música para a realização da tarefa. Enquanto a maioria dos trabalhos do segundo tipo utilizam-se do conteúdo rítmico e tímbrico, este utiliza-se apenas de conceitos da teoria da informação e da geometria de fractais. Entropia, lacunaridade e dimensão do fractal são os parâmetros que treinam os classificadores. Os testes foram realizados com duas coleções criadas para este trabalho e os resultados foram proeminentesThe goal of automatic music genre classification is givingmusic listeners ease and confort when managing digital music databases. Some systems are based on tags of metadata (such as artist name, genre labeled, etc.), while others explore characteristics from the music files to complete the task. While the majority of works of the second type analyse rhytmic, timbric and pitch content, this one explores only information theoretic and fractal geometry concepts. Entropy, fractal dimension and lacunarity are the parameters adopted to train the classifiers. Tests were carried out on two databases assembled by the author. Results were prominent
3
Loeffler, Dominik B. "Instrument Timbres and Pitch Estimation in Polyphonic Music." Thesis, Georgia Institute of Technology, 2006. http://hdl.handle.net/1853/10568.
Full textAbstract:
In the past decade, the availability of digitally encoded, downloadable music has increased dramatically, pushed mainly by the release of the now famous MP3 compression format (Fraunhofer-Gesellschaft, 1994). Online sales of music in the US doubled in 2005, according to a recent news article (*), while the number of files exchanged on P2P platforms is much higher, but hard to estimate.
The existing and coming informational flood in digital music prompts the need for sophisticated content-based information retrieval. Query-by-Humming is a prototypical technique aimed at locating pieces of music by melody; automatic annotation algorithms seek to enable finer search criteria, such as instruments, genre, or meter. Score transcription systems strive for an abstract, compressed form of a piece of music understandable by composers and musicians.
Much research still has to be performed to achieve these goals.
This thesis connects essential knowledge about music and human auditory perception with signal processing algorithms to solve the specific problem of pitch estimation. The designed algorithm obtains an estimate of the magnitude spectrum via STFT and models the harmonic structure of each pitch contained in the magnitude spectrum with Gaussian density mixtures, whose parameters are subsequently estimated via an Expectation-Maximization (EM) algorithm.
Heuristics for EM initialization are formulated mathematically.
The system is implemented in MATLAB, featuring a GUI that provides for visual (spectrogram) and numerical (console) verification of results. The algorithm is tested using an array of data ranging from single to triple superposed instrument recordings. Its advantages and limitations are discussed, and a brief outlook over potential future research is given.
(*) "Online and Wireless Music Sales Tripled in 2005"; Associated Press; January 19, 2006
4
Šolc, Radek. "Segmentace cévního řečiště na snímcích sítnice s využitím statistických metod." Master's thesis, Vysoké učení technické v Brně. Fakulta elektrotechniky a komunikačních technologií, 2015. http://www.nusl.cz/ntk/nusl-221344.
Full textAbstract:
This diploma thesis deals with segmentation of blood vessel from images acquired by fundus camera. The characteristic of fundus images and current methods of segmentation are described in theoretical part. The reach of the practical part is method using statistical model. The model using Student´s distribution for automatic segmentation is gradually drafted. Firstly EM- algorithm has been incorporated and model drafted on Markov random fields for improving robustness to noise after that. Contrast of thin blood vessel is improved in image preprocessing part by discrete wave transformation. The output image is used as mask for grayscale intensity decrease of thinnest blood-vessel and intensity increase of background. Whole model was programed in Matlab. The model was tested on whole HRF database. The quantitative evaluation of binary images were compared with golden standard images.
5
Vieira, Sylvio André Garcia. "IDENTIFICAÇÃO DE PADRÕES DE EXPRESSÃO EM DOENÇAS GENÉTICAS USANDO UMA REDE DE INTEGRAÇÃO DE VIAS DE MANUTENÇÃO DO GENOMA, ANGIOGÊNESE, HIPÓXIA EVIGILÂNCIA IMUNOLÓGICA." Centro Universitário Franciscano, 2016. http://www.tede.universidadefranciscana.edu.br:8080/handle/UFN-BDTD/571.
Full textAbstract:
Submitted by MARCIA ROVADOSCHI (marciar@unifra.br) on 2018-08-20T12:12:52Z
No. of bitstreams: 2
license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5)
Tese_SylvioAndreGarciaVieira.pdf: 3455615 bytes, checksum: 2826b01774c4732fb20d76dd1dde7a09 (MD5)Made available in DSpace on 2018-08-20T12:12:52Z (GMT). No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Tese_SylvioAndreGarciaVieira.pdf: 3455615 bytes, checksum: 2826b01774c4732fb20d76dd1dde7a09 (MD5) Previous issue date: 2016-08-23
The analysis of the biological formations through computer programs has been turning the interpretation of information quicker, more practical, and more reliable. There are a great number of data stored in specialized databases all over the world, that need to be analyzed and interpreted. On the biological databases, there are data obtained from a variety of ways of research in organisms affected by illnesses such as adenoma and carcinoma. There are also data related to the processes of maintenance of the organism in various levels, called pathways. These pathways act in the organism in a different way in every stage, such as in its normality or in the presence of some genetic disease. This study aims at using graphs in order to develop a model of pathway interaction network that incorporate the maintenance of the genome and other activities of the organism conveyed in adenoma and carcinoma of the adrenal cortex, identifying those who are active in the organism in the presence of gene mutation. In order to determine which pathways are modified in the presence of each disease was used a calculus of relative activity of the route associated to the statistical test Z. To visualize the results, graphs were used, whose junctions represent the pathways and whose edges represent the interactions. Through the computer program developed, it was possible to identify the differences that the organism presents in these conditions, allowing the suggestion of employment of this technique to identify modifications that the organism may present when a new nanoencapsulated drug is used.
A análise de informações biológicas por meios computacionais vem tornando as atividades de interpretação das informações mais rápida, prática e confiável. Existem muitos dados depositados em bancos especializados ao redor do mundo, que precisam ser avaliados e interpretados. Nestes bancos de dados de informações biológicas há informações resultantes de várias formas de pesquisas em organismos acometidos de doenças, como adenomas e carcinomas. A evolução dessas doenças ocorre através da ativação de vias muito específicas. Estas vias atuam no organismo de forma diferente em cada estado, como na sua normalidade ou na presença de alguma doença genética. Este trabalho tem por objetivo utilizar grafos para desenvolver um modelo de redes de interação de vias que envolvam a manutenção do genoma e outras atividades do organismo expressas em adenoma e carcinoma de córtex adrenal, identificando as vias expressas no organismo na presença de mutações. Para determinar quais vias estariam com a sua expressão alterada na presença de cada doença, foi utilizado o cálculo de atividade relativa da via associado ao teste estatístico Z. Para visualização dos resultados foram utilizados grafos, cujos nós representam as vias e as arestas representam as interações. Por meio do sistema computacional desenvolvido foi possível identificar as diferenças que o organismo apresenta nestas condições, permitido a sugestão de utilização desta técnica para a identificação das modificações que o organismo possa apresentar quando da utilização de um novo fármaco nanoencapsulado.
6
Rocha, Marcos Aurélio Andrade. "Ensaios sobre o papel da taxa de câmbio no desenvolvimento." reponame:Repositório Institucional do FGV, 2012. http://hdl.handle.net/10438/9859.
Full textAbstract:
Submitted by Marcos Rocha (marcosrocha@gvmail.br) on 2012-06-27T18:30:38Z
No. of bitstreams: 1
Tese_MRocha_biblio.pdf: 2007577 bytes, checksum: 595f726cc8507c10b02153b1ee6674cd (MD5)Approved for entry into archive by Vera Lúcia Mourão (vera.mourao@fgv.br) on 2012-06-27T19:02:57Z (GMT) No. of bitstreams: 1 Tese_MRocha_biblio.pdf: 2007577 bytes, checksum: 595f726cc8507c10b02153b1ee6674cd (MD5)
Made available in DSpace on 2012-06-27T19:11:03Z (GMT). No. of bitstreams: 1 Tese_MRocha_biblio.pdf: 2007577 bytes, checksum: 595f726cc8507c10b02153b1ee6674cd (MD5) Previous issue date: 2012-05-29
Este trabalho é composto de três partes. O primeiro capítulo avalia a hipótese de substituição de poupanças. De acordo com Bresser-Pereira e Nakano (2003), existe nas economias emergentes uma alta taxa de substituição de poupança interna por externa gerada pelo processo de influxo dos recursos estrangeiros. Isso acontece porque, dada a propensão ao consumo dos países emergentes, a maior parte do déficit em conta corrente se transforma em consumo de importados; a parcela dos fluxos gasta em investimento é pequena, de forma que a estratégia de crescimento com poupança estrangeira não é uma opção. Diante desta hipótese, o capítulo avalia teórica e empiricamente a existência de substituição de poupança interna por externa nos países em desenvolvimento e o papel da sobreapreciação da taxa de câmbio real (TCR) nesse processo. Foi utilizado um painel dinâmico System-GMM para uma amostra de 48 países de renda média cobrindo o período de 1970-2004. Os resultados mostram um processo de substituição de poupança doméstica por externa significante, além de efeitos negativos sobre a poupança doméstica resultantes da interação entre sobreapreciação do câmbio e entrada de poupança estrangeira. O segundo capítulo analisa os efeitos de sobreapreciações da TCR sobre o valor adicionado por setores dos países emergentes, para o período 1970-2004. A sobreapreciação, ao minar a competitividade dos produtos manufaturados e industriais domesticamente, pode reduzir a capacidade dos países em desenvolvimento de estimular setores dinâmicos da estrutura produtiva que induzem crescimento e desenvolvimento sustentável no longo prazo. A motivação teórica é a visão Kaldoriana de desenvolvimento. Assim, o capítulo testa a relação entre a sobreapreciação da TCR sobre o valor adicionado dos setores produtivos de bens primários e manufatureiros como parcela do produto, no período de 1970 a 2004, para uma amostra de 45 países emergentes. Os resultados evidenciam que os episódios de sobreapreciação caracterizados pelo índice reduzem a participação de manufaturados no valor agregado no produto das economias em desenvolvimento, de forma significante. A razão provável deste resultado é que um câmbio sobreapreciado não só torna os bens industriais domésticos menos competitivos, como também favorece a importação doméstica destes que, em virtude disto, deixam de ser produzidos ou aperfeiçoados dentro do país. Por fim, o terceiro capítulo avalia os efeitos da dívida externa sobre crescimento de um ponto de vista diferente dos diagnósticos da literatura 'debt overhang': argumenta que é a estratégia de crescimento com poupança externa a causa do crescente endividamento que, no longo prazo, é responsável por prejudicar a performance de crescimento dos países emergentes por gerar crises no balanço de pagamentos. Para avaliar esta dinâmica empiricamente foi estimada uma equação de crescimento com painel dinâmico System-GMM de variáveis instrumentais; são utilizados dados para 45 países de renda média, no período 1970-2004. A conclusão geral dos exercícios empíricos atesta uma influência não linear da dívida externa sobre o crescimento dos países selecionados na amostra: no longo prazo, o contínuo endividamento reduz o desempenho de crescimento dos países. Pode-se apontar ainda que um dos mecanismos pelo qual a poupança externa pode levar a maior dívida externa é sua interação significante com o influxo de recursos externos necessários para a rolagem da dívida; o resultado sobre o desempenho de crescimento é negativo e significante. Por fim, o apêndice do trabalho detalha os procedimentos de estimação de painel cointegrado utilizados para derivar os índices de sobreapreciação da TCR.
This work consists of three parts. The first chapter assesses the possibility of “savings displacement”. According to Bresser-Pereira and Nakano (2003), there is in emerging economies a high rate of substitution of foreign for domestic savings generated by the process of inflow of foreign financial flows. This is happens because, given the propensity for consumption in emerging countries, most of the deficit in current account turns into consumption of imported goods; the share of the flow spent on investment is small. The existence of savings displacement makes growing with foreign savings a risky strategy. Given this hypothesis, the chapter evaluates theoretically and empirically the existence of substitution of foreign for domestic savings in developing countries and also the role of overvaluation of the real exchange rate (RER) in this process. We used a dynamic panel GMM-System for a sample of 48 middle income countries covering the time span of 1970-2004. The results show a process of substitution of foreign for domestic savings significant and negative effects on domestic savings resulting from the interaction of overvaluation of the exchange and foreign inflows. The second chapter analyzes the effects of RER overvaluations on the value added by economy sectors for developing countries covering the period 1970-2004. The overvaluation process, once undermines the competitiveness of domestically manufacturing industries may reduce the ability of developing countries to stimulate dynamic sectors of the productive structure. The long run outcome can led to adverse effects upon growth and sustainable development. The theoretical motivation of this work is the Kaldorian view of development. The chapter proceeds by testing tests the relationship between the overvaluation of the RER on the added value of the productive sectors of primary goods and manufacturing as product share for the period 1970 to 2004, using a sample of 45 emerging countries. The results show that episodes of overvaluation are correlated with a lower share of manufacturing value-added production on the economies´ product significantly. The probable reason of this result is that an exchange overvaluation not only makes the domestic industrial goods less competitive, but also favors the domestic importats of this commodity from other countries; in this scenario, manufacturing could no longer be produced or improved within the country. Finally, the third chapter evaluates the effects of external debt on growth from a different point of view underlined by the "debt overhang" literature. Here we argues that the strategy of growth with foreign savings is the main reason for the increasing debt that in the long term that is responsible for harming the growth performance of developing countries since generates balance of payments crises. To assess this dynamic was empirically estimated growth equation with a dynamic panel System GMM. The data used 45 middle-income countries, covering the period 1970-2004. The overall conclusion of the empirical exercises attests a nonlinear influence of external debt on growth of selected countries in the sample: in the long term, continuous indebtedness reduces growth performance of countries. It can be pointed that one probable reason why foreign savings can lead to higher external debt problems is its significant interaction with the influx of external resources required to rollover debt, the result on the growth performance is negative and significant. Concluding the thesis, an appendix details the procedures for estimating panel cointegration used to derive indexes of RER overvaluation.
7
Costa, Paulo Renato Freitas. "Aproximações assimptóticas na análise da heterogeneidade negligenciada em modelos de duração." Master's thesis, Instituto Superior de Economia e Gestão, 2005. http://hdl.handle.net/10400.5/18849.
Full textAbstract:
Mestrado em Econometria Aplicada e PrevisãoNa análise econométrica de modelos de duração, é bastante comum depararmo- nos com a existência de heterogeneidade negligenciada. Este problema deve-se, em grande parte, ao facto de não ser possível observar todas as características individuais que afectam a duração de um evento. Neste trabalho, utilizando aproximações assimptóticas, vamos estudar o im- pacto da heterogeneidade em modelos de duração com parâmetro de escala. Como as aproximações são sensíveis ao efeito da heterogeneidade, irão constituir a base para a construção de um teste para a sua detecção. Uma expressão que caracterize o en- viesamento causado pela heterogeneidade será derivada utilizando as aproximações, sendo essa expressão a base para a construção de um estimador GMM que corrige o enviesamento na condição de momentos do modelo que ignora a heterogeneidade. Utilizando duas distribuições paramétricas, Weibull e Log-logistic, e através de simulações de Monte Carlo, iremos analisar o comportamento do estimador GMM proposto.
In the econometric analysis of duration models, it is common to come across itself with the problem of neglected heterogeneity. This problem, to a large extent, is due to the fact of not being possible to observe ali the individual characteristics that affect the duration of an event. In this work, small parameters aproximations will be used to study the impact of the heterogeneity in duration models with a scale paramter. The aproximations are sensible to the effect of the heterogeneity so, they will be the basis to construct a test for its detection. An expression that characterizes the bias caused by neglecting heterogeneity will be derived using the aproximations, being this expression the base for the constrution of a GMM that corrects the bias in the moment conditions of the model that ignores heterogeneity. Using two parametric distributions, Weibull and Log-logistic, and through Monte Carlo simulation, we will analyze the perfomance of the proposed GMM estimator.
info:eu-repo/semantics/publishedVersion
8
Campos, Victor de Abreu [UNESP]. "Arcabouço para reconhecimento de locutor baseado em aprendizado não supervisionado." Universidade Estadual Paulista (UNESP), 2017. http://hdl.handle.net/11449/151725.
Full textAbstract:
Submitted by Victor de Abreu Campos null (victorde.ac@gmail.com) on 2017-09-27T02:41:28Z
No. of bitstreams: 1
dissertacao.pdf: 5473435 bytes, checksum: 1e76ecc15a4499dc141983740cc79e5a (MD5)Approved for entry into archive by Monique Sasaki (sayumi_sasaki@hotmail.com) on 2017-09-28T13:43:21Z (GMT) No. of bitstreams: 1 campos_va_me_sjrp.pdf: 5473435 bytes, checksum: 1e76ecc15a4499dc141983740cc79e5a (MD5)
Made available in DSpace on 2017-09-28T13:43:21Z (GMT). No. of bitstreams: 1 campos_va_me_sjrp.pdf: 5473435 bytes, checksum: 1e76ecc15a4499dc141983740cc79e5a (MD5) Previous issue date: 2017-08-31
Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)
A quantidade vertiginosa de conteúdo multimídia acumulada diariamente tem demandado o desenvolvimento de abordagens eficazes de recuperação. Nesse contexto, ferramentas de reconhecimento de locutor capazes de identificar automaticamente um indivíduo pela sua voz são de grande relevância. Este trabalho apresenta uma nova abordagem de reconhecimento de locutor modelado como um cenário de recuperação e usando algoritmos de aprendizado não supervisionado recentes. A abordagem proposta considera Coeficientes Cepstrais de Frequência Mel (MFCCs) e Coeficientes de Predição Linear Perceptual (PLPs) como características de locutor, em combinação com múltiplas abordagens de modelagem probabilística, especificamente Quantização Vetorial, Modelos por Mistura de Gaussianas e i-vectors, para calcular distâncias entre gravações de áudio. Em seguida, métodos de aprendizado não supervisionado baseados em ranqueamento são utilizados para aperfeiçoar a eficácia dos resultados de recuperação e, com a aplicação de um classificador de K-Vizinhos Mais Próximos, toma-se uma decisão quanto a identidade do locutor. Experimentos foram conduzidos considerando três conjuntos de dados públicos de diferentes cenários e carregando ruídos de diversas origens. Resultados da avaliação experimental demonstram que a abordagem proposta pode atingir resultados de eficácia altos. Adicionalmente, ganhos de eficácia relativos de até +318% foram obtidos pelo procedimento de aprendizado não supervisionado na tarefa de recuperação de locutor e ganhos de acurácia relativos de até +7,05% na tarefa de identificação entre gravações de domínios diferentes.
The huge amount of multimedia content accumulated daily has demanded the development of effective retrieval approaches. In this context, speaker recognition tools capable of automatically identifying a person through their voice are of great relevance. This work presents a novel speaker recognition approach modelled as a retrieval scenario and using recent unsupervised learning methods. The proposed approach considers Mel-Frequency Cepstral Coefficients (MFCCs) and Perceptual Linear Prediction Coefficients (PLPs) as features along with multiple modelling approaches, namely Vector Quantization, Gaussian Mixture Models and i-vector to compute distances among audio objects. Next, rank-based unsupervised learning methods are used for improving the effectiveness of retrieval results and, based on a K-Nearest Neighbors classifier, an identity decision is taken. Several experiments were conducted considering three public datasets from different scenarios, carrying noise from various sources. Experimental results demonstrate that the proposed approach can achieve very high effectiveness results. In addition, effectiveness gains up to +318% were obtained by the unsupervised learning procedure in a speaker retrieval task. Also, accuracy gains up to +7,05% were obtained by the unsupervised learning procedure in a speaker identification task considering recordings from different domains.
FAPESP: 2015/07934-4
9
Moraes, Flávio Luiz Alves Flores. "Estimação de parâmetros de demanda e oferta em mercados de produtos diferenciados." reponame:Repositório Institucional do FGV, 2008. http://hdl.handle.net/10438/6963.
Full textAbstract:
Submitted by FLAVIO MORAES (flafmoraes@yahoo.com.br) on 2010-08-19T14:21:41Z
No. of bitstreams: 1
Dissertacao_Flavio_Moraes.pdf: 462004 bytes, checksum: 123425d1cbdd74b47091293eaf1f296d (MD5)Approved for entry into archive by Andrea Virginio Machado(andrea.machado@fgv.br) on 2010-08-19T16:53:50Z (GMT) No. of bitstreams: 1 Dissertacao_Flavio_Moraes.pdf: 462004 bytes, checksum: 123425d1cbdd74b47091293eaf1f296d (MD5)
Made available in DSpace on 2010-08-20T14:46:06Z (GMT). No. of bitstreams: 1 Dissertacao_Flavio_Moraes.pdf: 462004 bytes, checksum: 123425d1cbdd74b47091293eaf1f296d (MD5) Previous issue date: 2008-07-18
The purpose of this paper is to discuss in details the development of estimation methods of demand and supply parameters in di§erentiated products markets. The techniques presented consider explicitly the endogeneity of prices and can be applied to di§erent types of industries. The market demands system is derived from discrete choice models describing the behaviour of the consumer. This system is then combined with hypothesis about the cost functions and the behaviour of price determination by the Örms to generate equilibrium prices and quantities. The parameters to be estimated are those that determine the marginal costs of the Örms and the distribution of the consumersí tastes. This distribution determine elasticities and these, combined with the marginal cost and a hypothesis of Nash equilibrium, determine equilibrium prices. These elasticities and cost parameters play a central role in analysis of descriptive
Este trabalho tem por objetivo discutir detalhadamente o desenvolvimento de métodos de estimação de parâmetros de demanda e oferta em mercados de produtos diferenciados. As técnicas apresentadas consideram explicitamente a endogeneidade dos preços e podem ser aplicadas a diferentes tipos de indústrias. O sistema de demandas de mercado é derivado a partir de modelos de escolha discreta descrevendo o comportamento do consumidor. Esse sistema é então combinado com hipóteses sobre as funções custo e sobre o comportamento de determinação dos preços por parte das firmas para gerar preços e quantidades de equilíbrio. Os parâmetros a ser estimados são os que determinam os custos marginais das firmas e a distribuição dos gostos dos consumidores. Essa distribuição determina elasticidades e estas, combinadas com o custo marginal e com uma hipótese de equilíbrio de Nash na determinação de preços, determinam preços de equilíbrio. Essas elasticidades e parâmetros de custo desempenham um papel central em análises de questões descritivas e de mudanças no ambiente do mercado sob análise.
10
Gomes, Barbara Marisa Moreira. "Consumo energético e crescimento económico: que relação é possível estabelecer? Análise para um painel de países da UE." Master's thesis, 2017. http://hdl.handle.net/10316/82718.
Full textAbstract:
Trabalho de Projeto do Mestrado em Economia apresentado à Faculdade de EconomiaTomando em linha de conta a crescente importância do consumo de energia verificado nas últimas décadas, torna-se relevante averiguar se é possível estabelecer uma ligação entre esse consumo crescente e o crescimento económico, principalmente no âmbito de economias desenvolvidas. Este estudo prende-se com o objetivo de analisar a relação entre consumo de energia e crescimento económico tendo em conta outros fatores, como o nível inicial do produto per capita, o capital físico e humano e o comércio externo. Analisa-se a importância que estas variáveis têm no crescimento económico em catorze países da União Europeia (UE), para o período compreendido entre 1986 a 2013, estimando equações de convergência em painel pelo método GMM (Método dos Momentos Generalizados), dada a presença de endogeneidade das variáveis explicativas. Para além das variáveis habitualmente usadas e consideradas nos modelos de crescimento económico de base neoclássica, acrescentamos, para além de proxies de capital humano e comércio externo, aquela que para nós é a variável de interesse, o consumo energético (que representamos por duas proxies alternativas, consumo de energia primária e emissões de dióxido de carbono, de acordo com a literatura sobre o tema).Em termos gerais, concluímos pela existência de convergência condicional, que o capital físico e o comércio externo têm o tipo de impacto esperado de acordo com a teoria económica; quanto ao capital humano e consumo de energia, os resultados podem ser considerados inesperados à partida, embora igualmente justificáveis.
Taking into account the growing importance of energy consumption in the last decades, it is relevant to ascertain whether it is possible to establish a link between this growing consumption and economic growth, especially in the context of developed economies. This study aims at analyzing the relationship between energy consumption and economic growth, taking into account other factors, such as the initial per capita product level, physical and human capital and external trade. It is analyzed the importance of these variables for economic growth in fourteen European Union (EU) countries, for the period 1986 to 2013, by estimating panel convergence equations by the GMM (Generalized Moments Method) method, given the presence of endogeneity of the explanatory variables. In addition to the variables commonly used and considered in neoclassical economic growth models, we have added, in addition to human capital and foreign trade proxies, our variable of interest, the energy consumption (which we represent by two alternative proxies, primary energy consumption and carbon dioxide emissions, according to the literature on the subject).In general terms, we observe the existence of conditional convergence, that physical capital and external trade have expected kind of impact expected according to economic theory. With regard to human capital and energy consumption, the results can be considered to be unexpected at first, although justifiable. .
Book chapters on the topic "GMM-EM"
1
Zablotskiy, Sergey, Teerat Pitakrat, Kseniya Zablotskaya, and Wolfgang Minker. "GMM Parameter Estimation by Means of EM and Genetic Algorithms." In Human-Computer Interaction. Design and Development Approaches, 527–36. Berlin, Heidelberg: Springer Berlin Heidelberg, 2011. http://dx.doi.org/10.1007/978-3-642-21602-2_57.
Full text
2
Bhavya, D., D. S. Vinod, S. P. Shiva Prakash, and Kiril Krinkin. "GMM-EM-ACO Model for Congestion Free Routing in Social Internet of Things." In Lecture Notes in Electrical Engineering, 295–303. Singapore: Springer Nature Singapore, 2023. http://dx.doi.org/10.1007/978-981-19-6581-4_23.
Full text
3
Huang, Weizhuo, Yinwei Zhan, and Rongqian Yang. "A Hybrid Model for Liver Shape Segmentation with Customized Fast Marching and Improved GMM-EM." In Lecture Notes in Computer Science, 495–508. Cham: Springer International Publishing, 2019. http://dx.doi.org/10.1007/978-3-030-34110-7_41.
Full text
4
Khoulqi, Ichrak, Najlae Idrissi, and Muhammad Sarfraz. "Segmentation of Pectoral Muscle in Mammogram Images Using Gaussian Mixture Model-Expectation Maximization." In Research Anthology on Medical Informatics in Breast and Cervical Cancer, 722–38. IGI Global, 2022. http://dx.doi.org/10.4018/978-1-6684-7136-4.ch038.
Full textAbstract:
Breast cancer is one of the significant issues in medical sciences today. Specifically, women are suffering most worldwide. Early diagnosis can result to control the growth of the tumor. However, there is a need of high precision of diagnosis for right treatment. This chapter contributes toward an achievement of a computer-aided diagnosis (CAD) system. It deals with mammographic images and enhances their quality. Then, the enhanced images are segmented for pectoral muscle (PM) in the Medio-Lateral-Oblique (MLO) view of the mammographic images. The segmentation approach uses the tool of Gaussian Mixture Model-Expectation Maximization (GMM-EM). A standard database of Mini-MIAS with 322 images has been utilized for the implementation and experimentation of the proposed technique. The metrics of structural similarity measure and DICE coefficient have been utilized to verify the quality of segmentation based on the ground truth. The proposed technique is quite robust and accurate, it supersedes various existing techniques when compared in the same context.
5
Khoulqi, Ichrak, Najlae Idrissi, and Muhammad Sarfraz. "Segmentation of Pectoral Muscle in Mammogram Images Using Gaussian Mixture Model-Expectation Maximization." In Advancements in Computer Vision Applications in Intelligent Systems and Multimedia Technologies, 162–77. IGI Global, 2020. http://dx.doi.org/10.4018/978-1-7998-4444-0.ch009.
Full textAbstract:
Breast cancer is one of the significant issues in medical sciences today. Specifically, women are suffering most worldwide. Early diagnosis can result to control the growth of the tumor. However, there is a need of high precision of diagnosis for right treatment. This chapter contributes toward an achievement of a computer-aided diagnosis (CAD) system. It deals with mammographic images and enhances their quality. Then, the enhanced images are segmented for pectoral muscle (PM) in the Medio-Lateral-Oblique (MLO) view of the mammographic images. The segmentation approach uses the tool of Gaussian Mixture Model-Expectation Maximization (GMM-EM). A standard database of Mini-MIAS with 322 images has been utilized for the implementation and experimentation of the proposed technique. The metrics of structural similarity measure and DICE coefficient have been utilized to verify the quality of segmentation based on the ground truth. The proposed technique is quite robust and accurate, it supersedes various existing techniques when compared in the same context.
6
Bhardwaj, Manish, Rekha Jain, Sumit Kumar, and Upma Kumari. "Use of Machine Learning Approaches in 5G Applications." In Advances in Wireless Technologies and Telecommunication, 137–49. IGI Global, 2023. http://dx.doi.org/10.4018/978-1-6684-7348-1.ch009.
Full textAbstract:
The rapid development of 5G technology has led to a proliferation of smart applications and the need for efficient network interoperability. To address the challenges associated with 5G applications, machine learning approaches have emerged as powerful tools for predicting network characteristics and optimizing resource allocation. This chapter explores the integration of advanced algorithmic techniques and distributed computing resources to enhance the performance of 5G networks. This chapter explores the integration of advanced algorithmic techniques and distributed computing resources to enhance the performance of 5G networks. It provides an overview of the integration of artificial intelligence (AI) with 5G technology, including logistic regression, support vector machines (SVM), artificial neural networks (ANN) such as MLPs, and clustering techniques like K-means clustering, Gaussian mixture model (GMM), and expectation maximization (EM). It also investigates the integration of 5G with existing mobile networks, emphasizing the importance of seamless connectivity and efficient handover mechanisms. Overall, this chapter provides a comprehensive analysis of the use of machine learning approaches in 5G applications. It highlights the potential of these techniques to address the unique challenges posed by 5G networks and enable the realization of their full potential. By leveraging advanced algorithmic techniques and integrating AI, network operators and service providers can optimize network performance, improve user experience, and unlock new opportunities in the era of 5G technology.
Conference papers on the topic "GMM-EM"
1
Levy, Tal, and Itsik Lapidot. "Selective GMM EM for telephone diarization." In 2016 IEEE International Conference on the Science of Electrical Engineering (ICSEE). IEEE, 2016. http://dx.doi.org/10.1109/icsee.2016.7806055.
Full text
2
Watanabe, Hidenori, Shogo Muramatsu, and Hisakazu Kikuchi. "Interval calculation of EM algorithm for GMM parameter estimation." In 2010 IEEE International Symposium on Circuits and Systems - ISCAS 2010. IEEE, 2010. http://dx.doi.org/10.1109/iscas.2010.5537044.
Full text
3
Zhao, Wentian, Shaojie Wang, Zhihuai Xie, Jing Shi, and Chenliang Xu. "GAN-EM: GAN Based EM Learning Framework." In Twenty-Eighth International Joint Conference on Artificial Intelligence {IJCAI-19}. California: International Joint Conferences on Artificial Intelligence Organization, 2019. http://dx.doi.org/10.24963/ijcai.2019/612.
Full textAbstract:
Expectation maximization (EM) algorithm is to find maximum likelihood solution for models having latent variables. A typical example is Gaussian Mixture Model (GMM) which requires Gaussian assumption, however, natural images are highly non-Gaussian so that GMM cannot be applied to perform image clustering task on pixel space. To overcome such limitation, we propose a GAN based EM learning framework that can maximize the likelihood of images and estimate the latent variables. We call this model GAN-EM, which is a framework for image clustering, semi-supervised classification and dimensionality reduction. In M-step, we design a novel loss function for discriminator of GAN to perform maximum likelihood estimation (MLE) on data with soft class label assignments. Specifically, a conditional generator captures data distribution for K classes, and a discriminator tells whether a sample is real or fake for each class. Since our model is unsupervised, the class label of real data is regarded as latent variable, which is estimated by an additional network (E-net) in E-step. The proposed GAN-EM achieves state-of-the-art clustering and semi-supervised classification results on MNIST, SVHN and CelebA, as well as comparable quality of generated images to other recently developed generative models.
4
Zhou, Xi-Yu, and Joon S. Lim. "EM algorithm with GMM and Naive Bayesian to Implement Missing Values." In Mobile and Wireless 2014. Science & Engineering Research Support soCiety, 2014. http://dx.doi.org/10.14257/astl.2014.46.01.
Full text
5
Medeiros, Marcus, Gabriel Araujo, Hendrik Macedo, Marco Chella, and Leonardo Matos. "Multi-kernel approach to Parallelization of EM Algorithm for GMM Training." In 2014 Brazilian Conference on Intelligent Systems (BRACIS). IEEE, 2014. http://dx.doi.org/10.1109/bracis.2014.38.
Full text
6
Markov, Konstantin P., and Seiichi Nakagawa. "Discriminative training of GMM using a modified EM algorithm for speaker recognition." In 5th International Conference on Spoken Language Processing (ICSLP 1998). ISCA: ISCA, 1998. http://dx.doi.org/10.21437/icslp.1998-224.
Full text
7
Jannah, Wardatul, and Dewi R. S. Saputro. "Parameter estimation of Gaussian mixture models (GMM) with expectation maximization (EM) algorithm." In INTERNATIONAL CONFERENCE OF MATHEMATICS AND MATHEMATICS EDUCATION (I-CMME) 2021. AIP Publishing, 2022. http://dx.doi.org/10.1063/5.0117119.
Full text
8
Liu, Ziquan, Lei Yu, Janet H. Hsiao, and Antoni B. Chan. "Parametric Manifold Learning of Gaussian Mixture Models." In Twenty-Eighth International Joint Conference on Artificial Intelligence {IJCAI-19}. California: International Joint Conferences on Artificial Intelligence Organization, 2019. http://dx.doi.org/10.24963/ijcai.2019/426.
Full textAbstract:
The Gaussian Mixture Model (GMM) is among the most widely used parametric probability distributions for representing data. However, it is complicated to analyze the relationship among GMMs since they lie on a high-dimensional manifold. Previous works either perform clustering of GMMs, which learns a limited discrete latent representation, or kernel-based embedding of GMMs, which is not interpretable due to difficulty in computing the inverse mapping. In this paper, we propose Parametric Manifold Learning of GMMs (PML-GMM), which learns a parametric mapping from a low-dimensional latent space to a high-dimensional GMM manifold. Similar to PCA, the proposed mapping is parameterized by the principal axes for the component weights, means, and covariances, which are optimized to minimize the reconstruction loss measured using Kullback-Leibler divergence (KLD). As the KLD between two GMMs is intractable, we approximate the objective function by a variational upper bound, which is optimized by an EM-style algorithm. Moreover, We derive an efficient solver by alternating optimization of subproblems and exploit Monte Carlo sampling to escape from local minima. We demonstrate the effectiveness of PML-GMM through experiments on synthetic, eye-fixation, flow cytometry, and social check-in data.
9
Papadimitriou, Dimitrios I., Zissimos P. Mourelatos, and Zhen Hu. "Reliability Analysis Using Second-Order Saddlepoint Approximation and Mixture Distributions." In ASME 2018 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference. American Society of Mechanical Engineers, 2018. http://dx.doi.org/10.1115/detc2018-85267.
Full textAbstract:
This paper proposes a new second-order Saddlepoint Approximation (SOSA) method for reliability analysis of nonlinear systems with correlated non-Gaussian and multimodal random variables. The proposed method overcomes the limitation of current available SOSA methods which are applicable to problems with only Gaussian random variables, by employing a Gaussian Mixture Model (GMM). The latter is first constructed using the Expectation Maximization (EM) method to approximate the joint probability density function of the input variables. Expressions of the statistical moments of the response variables are then derived using a second-order Taylor expansion of the limit-state function and the GMM. The standard SOSA method is finally integrated with the GMM to effectively analyze the reliability of systems with correlated non-Gaussian random variables. The accuracy of the proposed method is compared with existing methods including a SOSA based on Nataf transformation. Numerical examples demonstrate the effectiveness of the proposed approach.
10
Rodrigues, Ícaro De Lima, Davyd B. De Melo, Breno M. Freitas, and Danielo G. Gomes. "Detecção de Anomalias em Padrões Acústicos, de Temperatura e Umidade Sazonais para Abelhas Melíferas (Apis mellifera L.)." In Workshop de Computação Aplicada à Gestão do Meio Ambiente e Recursos Naturais. Sociedade Brasileira de Computação, 2021. http://dx.doi.org/10.5753/wcama.2021.15738.
Full textAbstract:
O Distúrbio do Colapso das Colônias (Colony Collapse Disorder, CCD) é um fenômeno associado ao desaparecimento repentino de abelhas melíferas em colônias manejadas. Registrado nos EUA desde 2006, o CCD tem como possíveis causas desde as variações climáticas, uso incorreto de defensivos químicos até pragas e doenças. Neste sentido, soluções computacionais criativas podem contribuir para um melhor entendimento da sanidade e do bem-estar das abelhas. Neste artigo, aplicamos modelos de aprendizagem de máquina para detectar anomalias em padrões acústicos de abelhas melíferas africanizadas (Apis mellifera L.) e em padrões de temperatura e de umidade sazonais internas em colmeias com abelhas de raça europeia. Três modelos preditivos foram implementados: Modelo de Mistura de Gaussianas (GMM), Máquina de Aprendizado Extremo (ELM) e Máquina de Vetor de Suporte para uma classe (OC-SVM). Utilizamos datasets com dados de temperatura e umidade internas de duas colmeias de abelhas melíferas localizadas nas cidades de Bournemouth (Inglaterra) e Würtzburg (Alemanha) e um dataset com áudio de uma colônia melífera em Fortaleza-CE (Brasil). Para temperatura e umidade, os melhores resultados ocorreram para anomalias sazonais, nas quais o algoritmo ELM alcançou uma acurácia média de 92,6%. Para os áudios, destacamos o algoritmo GMM (acurácia média de 84,9%) na detecção de ausência da rainha.