Search

In this paper, we investigate the impact of optimal hyper-parameter configuration in relational topic models. The main goal is to validate the hypothesis that single-objective Bayesian Optimization (BO) can discover a hyper-parameter setting that leads a set of relational topic models to simultaneously ensure good prediction capabilities and significant topics from a qualitative perspective. Our research, as a result of a comparative analysis performed on 7 state-of-the-art models, 5 performance measures and 3 datasets, has highlighted three main findings: (1) the majority of relational topic models are not able to offer a good trade-off between classification capabilities and topic interpretability; (2) single-objective optimization of hyper-parameters, targeted on maximizing the F1-Measure, is able to create topics that are also optimal with respect to the Kullback Leibler divergence measure; (3) the Pareto frontiers across several performance metrics reveals that the most promising trade-off between the performance metrics can be obtained by Constrained Relational Topic Models.

Serious games and gamification to engage consumers in pro-environmental behaviours for energy efficiency are in their infancy. To date, despite growing interest and some initial attempts, their potential to engage consumers in energy efficiency behaviours has been understudied. To provide an overview of serious games and gamification to engage individuals in pro-environmental behaviours for energy efficiency, a computer-based search for relevant publications was performed in several databases. After applying the inclusion criteria and removing duplicates, 10 papers were included in this systematic review. Results showed that serious games and gamification have been used in three different areas related to energy efficiency: environmental education, consumption awareness, and pro-environmental behaviours. This review also showed that applied gaming interventions can be used in more than one of these three areas (comprehensive interventions). The main observation to be drawn from this review is that both serious games and gamification can foster energy-saving behaviours and vary widely in terms of type of games and of features that might be appealing and motivating.

This systematic review focuses on the literature evidence for residual ovarian function during treatment with hormonal contraceptives. We reviewed all papers which assessed residual ovarian activity during hormonal contraceptive use, using endocrine markers such as serum anti-Müllerian hormone (AMH) concentrations, FSH, LH, oestradiol, progesterone and sonographic markers such as antral follicle count (AFC), ovarian volume and vascular indices. We considered every type (oestroprogestin or only progestin) and dosage of hormonal contraceptive and every mode of administration (oral, vaginal ring, implant, transdermal patch). We performed an electronic database search for papers published from 1 January 1990 until 30 November 2015 using PubMed and MEDLINE. We pre-selected 113 studies and judged 48 studies suitable for the review. Most studies showed that follicular development continues during treatment with hormonal contraceptives, and that during treatment there is a reduction in serum concentrations of FSH, LH and oestradiol, and also a reduction in endometrial thickness, ovarian volume and the number and size of antral follicles. The ovarian reserve parameters, namely AFC and ovarian volume, are lower among users than among non-users of hormonal contraception; regarding the effect of hormonal contraception on AMH, there are still controversies in the literature.

An approach based on hydraulic simulation and machine learning is presented, aimed at improving leakage management via analytical leak localization and reducing time and costs for investigation and rehabilitation of the Water Distribution Network. Hydraulic simulation is used to run different leakage scenarios by introducing a leak on each pipe, in turn, and varying its severity. The approach has been validated on two WDNs: a Pressure Management Zone in Milan (Italy) and a District Metered Area in Timisoara (Romania), the two pilots of the EU-FP7-ICT project ICeWater, obtaining a high reliability (>90%) in localizing a wide set of simulated leaks.

The on-line assessment of structural health of aircraft fuselage panels and their remaining useful life is crucial both in military and civilian settings. This paper presents an application of a Support Vector Machines (SVM) classification framework aimed at improving the diagnosis task based on the strain values acquired through a monitoring sensor network deployed on the helicopter fuselage panels. More in details, diagnosis is usually defined as detecting a damage, identifying the specific component affected (i.e., bay or stringer) and then characterizing the damage in terms of center and size. Here, the first two steps are performed through the SVM classification framework while the last one is based on an Artificial Neural Network (ANN) hierarchy already presented in a previous authors’ work. The training dataset was built through Finite Elements Method (FEM) based simulation, able to simulate the behavior of any type of panel and damage according to specific parameters to set up; the result of FE simulation consists of the strain fields on different locations. As results, the proposed SVM classification framework permits to improve reliability of detection and characterization tasks respect to the previous approach entirely based on ANN hierarchies. Finally, the remaining useful life is estimated by using another ANN, different for damage on bay and stringer, able to predict the values of two parameters of the NASGRO equation which is used to estimate the damage propagation.

A completely data-driven, fully adaptive self-learning algorithm for water demand forecasting in the short-term and with hourly periodicity is proposed, according to the renewed interest generated by the availability of new technological solutions such as Automatic Metering Readers (AMR), a key enabler of the “Smart Water” paradigm. The approach is based on two sequential stages: at the first stage (time-series clustering) the daily water demand patterns (i.e., time-series of hourly data) are analysed to identify a limited set of typical behaviours. At the second stage Support Vector Machine regression is used to obtain one specific forecasting model (consisting of a regression model for each hour) for each cluster identified at the first stage. The approach has been validated on real data acquired by AMRs deployed on the Italian pilot site of ICeWater, computing the widely adopted error measure MAPE (Mean Absolute Percentage Error).

Resilience and vulnerability of networked infrastructures are strictly linked: while resilience is focused on a general evaluation of the robustness of the entire infrastructure, vulnerability is associated with a specific component, or set of components, to represent the possibility of being influenced by hazards/threats and the severity of the possible consequences. In the proposedframework,complex network theory is used to evaluate resilience and vulnerability of a WDN (structural connectivity)whilehydraulic simulation (EPANET) is adopted to estimate the potential stress on pipes according to the level of service of the WDN. The aim is to provide a decision support functionality to assist the definition of intervention planbyrankingpipes according to integrated information on vulnerability/resilience and hydraulic stress. The proposed framework, developed within the ICeWater project, has been evaluated on both benchmark and two real world WDNs (pilots in Timisoara and Milan), starting from their hydraulic software models.

This paper presents a computational framework performing, in two stages: urban water demand pattern characterization through time series clustering and reliable hourly water demand forecasting for the entire day based on Support Vector Machine (SVM) regression. An SVM regression model is trained for each cluster identified and for each hour of the day, taking the hourly water demand data acquired at the very first m hours of the day. The approach has been validated on a real case study that is the urban water demand of the Water Distribution Network (WDN) in Milan, managed by Metropolitana Milanese, one of the partner of the EU-FP7-ICT ICeWater project.

A static model for the parallel execution of logic programs on MIMD distributed memory systems is presented where a refutation is implemented through a process network returned by the compilation of the logic program. The model supports Restricted-AND, OR and stream parallelism and it is integrated with a set of static analyses to optimise the process network. Altogether, the processes interact according to a static data driven model with medium grain operators. Data flowing in the network is tagged to distinguish bindings belonging to the same refutation. A scheduling strategy to integrate low level scheduling and message flow control has been defined. Performance figures are presented.