-Structured lipids (SLs) are novel triacylglycerols obtained by changing the native fatty acid (FA) profiles or by the incorporation of a new desired FA in the acylglycerol backbone. These modified fats present important medical and functional properties for food applications. This work aimed to synthetize a MLM-type SL, which consists of triacylglycerols containing a medium-chain FA (M) at sn-1,3 positions and a long-chain FA (L) at sn-2 position, by acidolysis of pumpkin seed oil with capric acid, catalyzed by a commercial lipase preparation from Thermomyces lanuginosa (Lipozyme TL IM). Reactions were performed at 45°C, in solvent-free media, at 1:2 molar ratio (pumpkin seed oil:capric acid) and a fixed amount of immobilized lipase of 5%, 10%, 15% or 20%. Incorporations of C10:0 increased with time up to 31 h (29.9 ± 0.7 mol-%) when 5% lipase load was used. Significant differences were only observed between the results obtained with 5 and 20% of biocatalyst load. The subsequent experiment was carried out with 5% lipase load, at 45°C, 1:2 molar ratio and in the presence of n-hexane. The results showed slightly higher incorporation yields in the presence of solvent, namely at 48 h-reaction (34.7 ± 1.0 mol-%). However, since the structured lipids are to be used in food products, together with environmental and economic concerns, solvent-free systems are preferred. In this study, the synthesis of a MLM-type SL from pumpkin seed oil for food uses was well succeeded.Keywords: structured lipid / acidolysis / pumpkin seed oil / capric acid / lipozyme TL IM Résumé -Incorporation d'acide caprique dans de l'huile de pépin de citrouille par acidolyse catalysée par une lipase sn-1,3 régiosélective. Les lipides structurés sont de nouveaux triacylglycérols obtenus en changeant les profils des acides gras (AG) natifs ou par l'incorporation de nouveaux AG souhaités dans le squelette glycérol. Ces graisses modifiées présentent des propriétés médicales et fonctionnelles importantes pour des applications alimentaires. Ce travail vise à synthétiser un lipide structuré de type MLM, qui consiste en des triacylglycérols contenant un AG à chaîne moyenne (M) en positions sn-1,3 et un acide gras à chaîne longue (L) en position sn-2, par acidolyse de l'huile de graine de citrouille avec l'acide caprique, catalysée par une préparation commerciale de lipase de Thermomyces lanuginosa (Lipozyme TL IM). Les réactions ont été effectuées à 45°C, sans solvant, avec un rapport molaire de 1:2 (huile de graines de citrouille: acide caprique) et une quantité fixe de lipase immobilisée de 5 %, 10 %, 15 % ou 20 %. Avec une charge en lipase de 5 %, l'incorporation du C10:0 a augmenté avec le temps jusqu'à 31 h (29,9 ± 0,7 mol %). Des différences significatives ont été observées seulement entre les résultats obtenus avec 5 et 20 % de charge en biocatalyseur. L'expérience suivante a été réalisée avec une charge en lipase de 5 %, à 45°C, un rapport molaire 1:2 et en présence de n-hexane. Les résultats ont montré des rendements d'incorporation lég...

No abstract

The ability of managers to analyze large volumes of data is not enough to identify all relevant associations and necessary for the decision-making process. Make use of a classification model and clustering model can generate information that typically a manager could not create without the utilization of this technology. The aim of this work is to reach a classification model linking them to clusters, based on data from purchases made by customers through electronic media in an automated manner. Precisely this model presents a set of rules to assist in decision-making applicable to a sale of vehicles, parts and accessories. For the construction of this model, we applied a process of knowledge discovery in databases. In which classification techniques and clustering techniques was evaluated in an experiment regarding accuracy, interpretability and learned a model of the computing performance. Data mining has been used to find this classifier.

RESUMOA busca por soluções sustentáveis visando minimizar o impacto ao meio ambiente tornou-se prioridade mundial, visto que é do saber popular que a maioria das fontes atuais de energia são primeiramente esgotáveis e também prejudiciais ao ecossistema terrestre. O mercado de energia elétrica no Brasil cresce aproximadamente na ordem de 4,5% ao ano, e em 2012 atingimos o consumo de 448,1 TWh. Entre os anos de 2008 a 2012 a população brasileira cresceu, em média, 1,5% a.a., e o consumo per capita de energia elétrica apresentou crescimento médio de 4,7% a.a. no mesmo período. Tendo conhecimento dessas informações é possível minimizar os investimentos necessários para atender à demanda crescente por energia elétrica, focando em reduzir desperdícios, visto que em um estudo realizado com indústrias, comércios e setor público, constatouse que os índices de desperdício chegam a 15%, 30% e 45%, respectivamente [EPE, 2013]. Nos últimos anos, os projetos de equipamentos e instalações elétricas já vêm sendo desenvolvidos com o foco em eficiência energética, mas a eficiência isoladamente não garante o uso racional da energia elétrica. O comportamento dos usuários e seus hábitos de utilização dos equipamentos representa hoje a variável mais flexível para a obtenção de economia de energia. Essa variável não sendo reduzida, obriga a automatização de sistemas para compensar o mau comportamento humano para com seu planeta. O presente trabalho tem como objetivo a avaliação da redução no consumo de energia, através da instalação de sensores de presença.

No abstract