Using the Big Data generated by the Smart Home to improve energy efficiency management

A Smart Home is able to generate energy-related values such as electricity consumption, temperature, or luminosity without higher infrastructure requirements. The main aim of this research is to extract information from that raw data that could contribute to improving the energy efficiency management. This paper presents a system which, using different Machine Learning approaches to learn about the users’ consumption habits, is able to generate collaborative recommendations and consumption predictions that help the user to consume better, which will in turn improve the demand curve. Moreover, from consumption values, the system learns to identify devices, enabling the demand to be anticipated. Taking into account the fact that the amount of energy data is increasing in real-time, the use of Big Data techniques will be the key to handling all the operations and one of the more innovative features of the system.     

Buoyancy effect on the ultrafiltration of PEG 10,000. Visualization by digital holographic interferometry

Digital holographic interferometry has been used to visualize the buoyancy effects on dead‐end ultrafiltration of PEG 10,000. Sets of experiments have been carried out with the membrane in different orientations (90^°, 180^°), using a feed concentration of PEG 10,000 ranging 5–12.5 kg/m³. These results were compared with those obtained in previous research with the cell placed in its natural position (0^°). The interferometric fringe patterns obtained in this research were very different from those obtained previously. Whereas at the 0^° position, an increasing number of interferometric fringes appeared parallel to the membrane surface at 90 and 180^° positions, the number of interference fringes was much lesser and, with the 180^° orientation, slightly curved protuberances appeared after a short period of time that, like big drops, fell downward from the membrane surface. The consequence is a decrease of the polarization layer and an enhancement of the membrane performance. © 2012 American Institute of Chemical Engineers AIChE J, 2012     

Evaluation methods and strategies for the interactive use of classifiers

We consider the scenario in which an automatic classifier (previously built) is available. It is used to classify new instances but, in some cases, the classifier may request the intervention of a human (the oracle), who gives it the correct class. In this scenario, first it is necessary to study how the performance of the system should be evaluated, as it cannot be based solely on the predictive accuracy obtained by the classifier but it should also take into account the cost of the human intervention; second, studying the concrete circumstances under which the classifier decides to query the oracle is also important. In this paper we study these two questions and include also an experimental evaluation of the different proposed alternatives.     

Location strategy for a firm under competitive delivered prices

This paper addresses the location problem for an entering firm that will play a Bertrand game with other pre-existing firms in order to maximize its profit. Demand for a homogeneous product is price-sensitive and firms use delivered pricing. Under some specific conditions, it is shown that the aggregate profit of the pre-existing facilities increases as the number of facilities of the entrant increases. The problem is analyzed for a network and for a discrete location space, showing that in both cases it can be solved by a Mixed Integer Linear Programming (MILP) formulation. This formulation is used to solve an illustrative example in a variety of scenarios, where profits for both the entering firm and its competitors are analyzed. In some scenarios is seen that the aggregate profit of the pre-existing facilities increases in the number of facilities of the entrant. It is also shown that locations are stabler with respect to changes in number of new facilities than with respect to changes in demand function.