A Data-Driven Framework for Small Hydroelectric Plant Prognosis Using Tsfresh and Machine Learning Survival Models.

Maintenance in small hydroelectric plants (SHPs) is essential for securing the expansion of clean energy sources and supplying the energy estimated to be required for the coming years. Identifying failures in SHPs before they happen is crucial for allowing better management of asset maintenance, lowering operating costs, and enabling the expansion of renewable energy sources. Most fault prognosis models proposed thus far for hydroelectric generating units are based on signal decomposition and regression models. In the specific case of SHPs, there is a high occurrence of data being censored, since the operation is not consistently steady and can be repeatedly interrupted due to transmission problems or scarcity of water resources. To overcome this, we propose a two-step, data-driven framework for SHP prognosis based on time series feature engineering and survival modeling. We compared two different strategies for feature engineering: one using higher-order statistics and the other using the Tsfresh algorithm. We adjusted three machine learning survival models-CoxNet, survival random forests, and gradient boosting survival analysis-for estimating the concordance index of these approaches. The best model presented a significant concordance index of 77.44%. We further investigated and discussed the importance of the monitored sensors and the feature extraction aggregations. The kurtosis and variance were the most relevant aggregations in the higher-order statistics domain, while the fast Fourier transform and continuous wavelet transform were the most frequent transformations when using Tsfresh. The most important sensors were related to the temperature at several points, such as the bearing generator, oil hydraulic unit, and turbine radial bushing.

[Efficiency of public spending on primary health care in the municipalities of Rio de Janeiro, Brazil: robust scores and their determinants]. / Eficiência do gasto público em atenção primária em saúde nos municípios do Rio de Janeiro, Brasil: escores robustos e seus determinantes.

Health services, the access to which is a right of all citizens and the duty of the state, must be efficiently provided, as public resources are scarce and there is a growing demand called for by the population. In this context, the objective was to evaluate the efficiency of public spending on Primary Health Care in the municipalities of Rio de Janeiro (n = 70) in 2015. Initially, the jackstrap procedure was used to identify possible outliers, and Data Envelopment Analysis was then applied to calculate robust efficiency scores. In the second stage, the efficiency scores were regressed into non-discretionary variables using the Tobit method, that may affect the efficiency of municipalities. The results show that 63.9% of municipalities efficiently applied ABS resources, however, given the expenses incurred, ABS outputs could have been, on average, 6% higher. In addition, no evidence was found that wealthier, more socially developed and/or populous municipalities, were more (or less) efficient in public spending on ABS than not.

Os serviços de saúde, cujo acesso é um direito de todos e dever do Estado, devem ser prestados com eficiência, pois os recursos públicos são escassos e há uma crescente demanda por parte da população. Nesse contexto, o objetivo deste artigo é avaliar a eficiência dos gastos públicos em atenção básica em saúde (ABS) nos municípios do Rio de Janeiro (n = 70) em 2015. Para isso, primeiramente, usa-se o procedimento jackstrap para identificar possíveis outliers e aplica-se a técnica de análise envoltória de dados para calcular escores robustos de eficiência. Em seguida, esses escores são regredidos, por meio de um modelo tobit, em variáveis ​​não discricionárias que possam afetar a eficiência dos municípios. Os resultados encontrados revelam que 63,9% dos municípios avaliados aplicaram eficientemente os recursos em ABS, mas que, diante dos gastos efetuados, os outputs em ABS poderiam ter sido em média 6,0% maiores. Além disso, não foram encontradas evidências de que municípios mais ricos, socialmente desenvolvidos e/ou populosos tenham sido mais (ou menos) eficientes com relação ao gasto público em ABS do que aqueles não o são.

Os determinantes da eficiência na Atenção Primária à Saúde dos municípios paulistas: um modelo georreferenciado / The determinants of efficiency in Primary Health Care in São Paulo municipalities: a georeferenced model

Resumo Esta pesquisa teve por objetivo mensurar a eficiência dos gastos em Atenção Primária à Saúde (APS) dos municípios paulistas e avaliar sua correlação com a densidade demográfica e o tamanho populacional dos municípios, assim como com sua localização geográfica e seu PIB per capita. Realizou-se uma avaliação ex post, quantitativa e descritiva acerca da eficiência do gasto em APS para uma amostra dos municípios paulistas (n=509), considerando os dados do ano de 2015. A presente pesquisa fez uso de um modelo de Data Envelopment Analysis com Supereficiência para detectar e remover outliers. Por fim, verificou-se a magnitude da relação existente entre os escores de eficiência calculados e as variáveis não discricionárias. Verificou-se que a média dos escores de eficiência após a exclusão dos municípios supereficientes (outliers) tornou-se maior (IC99%: 0.127; 0.213). Em média, os municípios paulistas são moderadamente ineficientes no tocante aos gastos com APS (ϴ = 0.488). Destaca-se que os municípios situados na faixa Leste do Estado de SP tendem a ter maiores níveis de eficiência e que a eficiência é inversamente proporcional ao tamanho populacional. O SUS possui notado subfinanciamento, mas há espaço para melhorias na gestão dos gastos e aumento da eficiência no serviço prestado.

Abstract This research aimed to measure the efficiency of expenditures on primary health care in the municipalities of São Paulo and to evaluate their correlation with demographic density and population size of municipalities, as well as with their geographic location and GDP per capita. An ex-post, quantitative, and descriptive valuation of the efficiency of SPS spending was carried out for a sample of the municipalities of São Paulo (n = 509), considering the data for the year 2015. The present research made use of a Data Envelopment Analysis model with Superefficiency to detect and remove outliers. Finally, the magnitude of the relationship between calculated efficiency scores and non-discretionary variables was verified. The average of the efficiency scores after the exclusion of the outliers municipalities became greater (IC99%: 0.127; 0.213). On average, São Paulo municipalities are moderately inefficient in terms of spending on APS (ϴ = 0.488). The municipalities located in the East range of the State of São Paulo tend to have higher levels of efficiency and that the efficiency is inversely proportional to the population size. The SUS has noticed underfunding; however, there is room for improvements in the management of expenses and increase efficiency in the service provided.