RESUMO
Near Infrared (NIR) is a type of vibrational spectroscopy widely used in different areas to characterize substances. NIR datasets are comprised of absorbance measures on a range of wavelengths (λ). Typically noisy and correlated, the use of such datasets tend to compromise the performance of several statistical techniques; one way to overcome that is to select portions of the spectra in which wavelengths are more informative. In this paper we investigate the performance of the Random Forest (RF) classifier associated with several wavelength importance ranking approaches on the task of classifying product samples into categories, such as quality levels or authenticity. Our propositions are tested using six NIR datasets comprised of two or more classes of food and pharmaceutical products, as well as illegal drugs. Our proposed classification model, an integration of the χ2 ranking score and the RF classifier, substantially reduced the number of wavelengths in the dataset, while increasing the classification accuracy when compared to the use of complete datasets. Our propositions also presented good performance when compared to competing methods available in the literature.
Assuntos
Análise de Dados , HumanosRESUMO
Resumo Este artigo analisa o financiamento público da cultura no Brasil, a partir de um estudo longitudinal quantitativo dos projetos aprovados pelo Ministério da Cultura entre 2003 e 2015, com base teórica na categoria de renda monopolista oriunda da economia política. Os resultados evidenciaram que, mesmo com a modificação na concepção de cultura no plano governamental, há concentração de incentivadores e de proponentes que utilizam a Lei de Incentivo à Cultura (LIC) para a obtenção de recursos no fomento a projetos. Concluímos que a política pública de financiamento da cultura continua transferindo para o mercado a decisão de quais projetos serão financiados. O mercado, por sua vez, tende à geração de renda monopolista, por meio do uso do recurso público para a promoção de projetos culturais que interessam à imagem organizacional sob o monopólio de poucas produtoras e fundações.
Resumen Este artículo analiza la financiación pública de la cultura en Brasil, a partir de un estudio longitudinal cuantitativo de los proyectos aprobados por el Ministerio de Cultura entre 2003 y 2015, con el análisis teórico basado en la categoría de la renta de monopolio de la economía política. Los resultados mostraron que incluso con el cambio en el concepto de cultura a nivel de gobierno, aún hay concentración de incentivadores y proponentes que utilizan la Ley de Incentivo a la Cultura (LIC) para obtener recursos en el fomento de proyectos. Concluimos que la política pública de la financiación de la cultura sigue transfiriendo al mercado para decidir qué proyectos serán financiados. El mercado, a su vez, tiende a generar renta de monopolio, a través del uso de los recursos públicos para la promoción de proyectos culturales de interés para la imagen de la organización bajo el monopolio de los pocos productores y fundaciones.
Abstract This article aims to analyze the public funding of culture in Brazil by studying the projects approved by the Ministry of Culture between 2003 and 2015. A quantitative longitudinal study is undertaken and includes theorizing on the concept of 'monopoly rent' from the political economy. The results showed that even with the change in the concept of culture at a government level, there is a concentration of supporters and proponents who use the Law of Cultural Incentive (LIC) to obtain resources to enable cultural projects. We conclude that the public policies of financing culture continue to transfer the decision of which projects will be funded to the market. The market, in turn, tends to generate monopoly rent with the use of public resources for the promotion of cultural projects that are monopolized by few producers and foundations.