Abstract
For extremely large telescopes, adaptive optics will be required to correct the Earth’s turbulent atmosphere. The performance of tomographic adaptive optics is strongly dependent on the vertical distribution (profile) of this turbulence. An important way in which this manifests is the tomographic error, arising from imperfect measurement and reconstruction of the turbulent phase at altitude. Conventionally, a small number of reference profiles are used to obtain this error in simulation; however these profiles are not constructed to be representative in terms of tomographic error. It is therefore unknown whether these simulations are providing realistic performance estimates. Here, we employ analytical adaptive optics simulation that drastically reduces computation times to compute tomographic error for 10 691 measurements of the turbulence profile gathered by the Stereo-SCIDAR instrument at ESO Paranal. We assess for the first time the impact of the profile on tomographic error in a statistical manner. We find, in agreement with previous work, that the tomographic error is most directly linked with the distribution of turbulence into discrete, stratified layers. Reference profiles are found to provide mostly higher tomographic error than expected, which we attribute to the fact that these profiles are primarily composed of averages of many measurements resulting in unrealistic, continuous distributions of turbulence. We propose that a representative profile should be defined with respect to a particular system, and that as such simulations with a large statistical sample of profiles must be an important step in the design process.

Abstract

Abstract
We propose a framework to model the distribution of sequential data coming from a set of entities connected in a graph with a known topology. The method is based on a mixture of shared hidden Markov models (HMMs), which are jointly trained in order to exploit the knowledge of the graph structure and in such a way that the obtained mixtures tend to be sparse. Experiments in different application domains demonstrate the effectiveness and versatility of the method.

Abstract
Email Marketing is one of the most important traffic sources in Digital Marketing. It yields a high return on investment for the company and offers a cheap and fast way to reach existent or potential clients. Getting the recipients to open the email is the first step for a successful campaign. Thus, it is important to understand how marketers can improve the open rate of a marketing campaign. In this work, we analyze what are the main factors driving the open rate of financial email marketing campaigns. For that purpose, we develop a classification algorithm that can accurately predict if a campaign will be labeled as Successful or Failure. A campaign is classified as Successful if it has an open rate higher than the average, otherwise it is labeled as Failure. To achieve this, we have employed and evaluated three different classifiers. Our results showed that it is possible to predict the performance of a campaign with approximately 82% accuracy, by using the Random Forest algorithm and the redundant filter selection technique. With this model, marketers will have the chance to sooner correct potential problems in a campaign that could highly impact its revenue. Additionally, a text analysis of the subject line and preheader was performed to discover which keywords and keyword combinations trigger a higher open rate. The results obtained were then validated in a real setting through A/B testing.

Abstract

Abstract
This paper presents a data-driven spatial net-load forecasting model that is applied to the distribution network expansion problem. The model uses population census data with Information Theory-based Feature Selection to predict spatial adoption patterns of residential electric vehicle chargers and photovoltaic modules. Results are high-resolution maps (0.02 km(2)) that allow distribution network planners to forecast asymmetric changes in load patterns and assess resulting impacts on installed HV/MV substation transformers in distribution systems. A risk analysis routine identifies the investment that minimizes the maximum regret function for a 15-year planning horizon. One of the outcomes from this study shows that traditional approaches to allocate distributed energy resources in distribution networks underestimate the impact of adopting EV and PV on the grid. The comparison of different allocation methods with the presented diffusion model suggests that using conventional approaches might result in strong underinvestment in capacity expansion during early uptake and overinvestment in later diffusion stages.