Cross-Approximate Entropy parallel computation on GPUs for biomedical signal analysis. Application to MEG recordings

Cross-Approximate Entropy (Cross-ApEn) is a useful measure to quantify the statistical dissimilarity of two time series. In spite of the advantage of Cross-ApEn over its one-dimensional counterpart (Approximate Entropy), only a few studies have applied it to biomedical signals, mainly due to its high computational cost. In this paper, we propose a fast GPU-based implementation of the Cross-ApEn that makes feasible its use over a large amount of multidimensional data. The scheme followed is fully scalable, thus maximizes the use of the GPU despite of the number of neural signals being processed. The approach consists in processing many trials or epochs simultaneously, with independence of its origin. In the case of MEG data, these trials can proceed from different input channels or subjects. The proposed implementation achieves an average speedup greater than 250× against a CPU parallel version running on a processor containing six cores. A dataset of 30 subjects containing 148 MEG channels (49 epochs of 1024 samples per channel) can be analyzed using our development in about 30 min. The same processing takes 5 days on six cores and 15 days when running on a single core. The speedup is much larger if compared to a basic sequential Matlab^® implementation, that would need 58 days per subject. To our knowledge, this is the first contribution of Cross-ApEn measure computation using GPUs. This study demonstrates that this hardware is, to the day, the best option for the signal processing of biomedical data with Cross-ApEn.     

Evolutionary discriminative confidence estimation for spoken term detection

Spoken term detection (STD) is the task of searching for occurrences of spoken terms in audio archives. It relies on robust confidence estimation to make a hit/false alarm (FA) decision. In order to optimize the decision in terms of the STD evaluation metric, the confidence has to be discriminative. Multi-layer perceptrons (MLPs) and support vector machines (SVMs) exhibit good performance in producing discriminative confidence; however they are severely limited by the continuous objective functions, and are therefore less capable of dealing with complex decision tasks. This leads to a substantial performance reduction when measuring detection of out-of-vocabulary (OOV) terms, where the high diversity in term properties usually leads to a complicated decision boundary. In this paper we present a new discriminative confidence estimation approach based on evolutionary discriminant analysis (EDA). Unlike MLPs and SVMs, EDA uses the classification error as its objective function, resulting in a model optimized towards the evaluation metric. In addition, EDA combines heterogeneous projection functions and classification strategies in decision making, leading to a highly flexible classifier that is capable of dealing with complex decision tasks. Finally, the evolutionary strategy of EDA reduces the risk of local minima. We tested the EDA-based confidence with a state-of-the-art phoneme-based STD system on an English meeting domain corpus, which employs a phoneme speech recognition system to produce lattices within which the phoneme sequences corresponding to the enquiry terms are searched. The test corpora comprise 11 h of speech data recorded with individual head-mounted microphones from 30 meetings carried out at several institutes including ICSI; NIST; ISL; LDC; the Virginia Polytechnic Institute and State University; and the University of Edinburgh. The experimental results demonstrate that EDA considerably outperforms MLPs and SVMs on both classification and confidence measurement in STD, and the advantage is found to be more significant on OOV terms than on in-vocabulary (INV) terms. In terms of classification performance, EDA achieved an equal error rate (EER) of 11% on OOV terms, compared to 34% and 31% with MLPs and SVMs respectively; for INV terms, an EER of 15% was obtained with EDA compared to 17% obtained with MLPs and SVMs. In terms of STD performance for OOV terms, EDA presented a significant relative improvement of 1.4% and 2.5% in terms of average term-weighted value (ATWV) over MLPs and SVMs respectively.     

Analysis of Secure Mobile Grid Systems: A systematic approach

Developing software through systematic processes is becoming more and more important due to the growing complexity of software development. It is important that the development process used integrates security aspects from the first stages at the same level as other functional and non-functional requirements. Systems which are based on Grid Computing are a kind of systems that have clear differentiating features in which security is a highly important aspect. The Mobile Grid, which is relevant to both Grid and Mobile Computing, is a full inheritor of the Grid with the additional feature that it supports mobile users and resources. A development methodology for Secure Mobile Grid Systems is proposed in which the security aspects are considered from the first stages of the life-cycle and in which the mobile Grid technological environment is always present in each activity. This paper presents the analysis activity, in which the requirements (focusing on the grid, mobile and security requirements) of the system are specified and which is driven by reusable use cases through which the requirements and needs of these systems can be defined. These use cases have been defined through a UML-extension for security use cases and Grid use cases which capture the behaviour of this kind of systems. The analysis activity has been applied to a real case.     

IT-enabled knowledge ambidexterity and innovation performance in small U.S. firms: The moderator role of social media capability

This study examines the impact of information technology (IT)-enabled knowledge ambidexterity on innovation performance, and the potential moderator role of social media capability on a sample composed of 100 small U.S. firms. The empirical analysis suggests that IT infrastructure enables the firm to explore new knowledge and exploit existing/new knowledge to innovate more and better. We also find that social media capability has a positive moderator role in this equation: IT infrastructure and social media capabilities work together to enable knowledge ambidexterity.     

A mixed–primal finite element method for the Boussinesq problem with temperature-dependent viscosity

In this paper we focus on the analysis of a mixed finite element method for a class of natural convection problems in two dimensions. More precisely, we consider a system based on the coupling of the steady-state equations of momentum (Navier–Stokes) and thermal energy by means of the Boussinesq approximation (coined the Boussinesq problem), where we also take into account a temperature dependence of the viscosity of the fluid. The construction of this finite element method begins with the introduction of the pseudostress and vorticity tensors, and a mixed formulation for the momentum equations, which is augmented with Galerkin-type terms, in order to deal with the non-linearity of these equations and the convective term in the energy equation, where a primal formulation is considered. The prescribed temperature on the boundary becomes an essential condition, which is weakly imposed, leading us to the definition of the normal heat flux through the boundary as a Lagrange multiplier. We show that this highly coupled problem can be uncoupled and analysed as a fixed-point problem, where Banach and Brouwer theorems will help us to provide sufficient conditions to ensure well-posedness of the problems arising from the continuous and discrete formulations, along with several applications of continuous injections guaranteed by the Rellich–Kondrachov theorem. Finally, we show some numerical results to illustrate the performance of this finite element method, as well as to prove the associated rates of convergence.     

Automatic collection of user behavior in 360° multimedia

The arrival of 360° video to the everyday life creates the necessity of assessing both the audiovisual production and the playback environment offered to the final user. Leveraging the standard Experience API (xAPI), that considers collecting micro-interactions with e-learning content, we propose a platform to automatically collect the users’ interaction with applications based on interactive 360° multimedia. To validate the platform, we introduce an example of educational activities based on interactive 360° videos and the tools used to first, annotate these videos and convert them into interactive activities; second, to perform said activity and collect the users’ behavior via xAPI statements; and finally, to convert these statements to meaningful information in the form of user metrics and charts, both at individual level and also aggregated by activity, creating the possibility of finding singular and group behavior. This work concludes that the presented platform helps to analyze how users behave with omnidirectional interactive productions, with the aim of validating and improving its usability, ending with the discussion of future work ideas.     

Topological Semantic Mapping and Localization in Urban Road Scenarios

Autonomous vehicle self-localization must be robust to environment changes, such as dynamic objects, variable illumination, and atmospheric conditions. Topological maps provide a concise representation of the world by only keeping information about relevant places, being robust to environment changes. On the other hand, semantic maps correspond to a high level representation of the environment that includes labels associated with relevant objects and places. Hence, the use of a topological map based on semantic information represents a robust and efficient solution for large-scale outdoor scenes for autonomous vehicles and Advanced Driver Assistance Systems (ADAS). In this work, a novel topological semantic mapping and localization methodology for large-scale outdoor scenarios for autonomous driving and ADAS applications is presented. The methodology uses: (i) a deep neural network for obtaining semantic observations of the environment, (ii) a Topological Semantic Map (TSM) for storing selected semantic observations, and (iii) a topological localization algorithm which uses a Particle Filter for obtaining the vehicle’s pose in the TSM. The proposed methodology was tested on a real driving scenario, where a True Estimate Rate of the vehicle’s pose of 96.9% and a Mean Position Accuracy of 7.7[m] were obtained. These results are much better than the ones obtained by other two methods used for comparative purposes. Experiments also show that the method is able to obtain the pose of the vehicle when its initial pose is unknown.     

A reactive simheuristic using online data for a real‐life inventory routing problem with stochastic demands

In the context of a supply chain for the animal‐feed industry, this paper focuses on optimizing replenishment strategies for silos in multiple farms. Assuming that a supply chain is essentially a value chain, our work aims at narrowing this chasm and putting analytics into practice by identifying and quantifying improvements on specific stages of an animal‐feed supply chain. Motivated by a real‐life case, the paper analyses a rich multi‐period inventory routing problem with homogeneous fleet, stochastic demands, and maximum route length. After describing the problem and reviewing the related literature, we introduce a reactive heuristic, which is then extended into a biased‐randomized simheuristic. Our reactive approach is validated and tested using a series of adapted instances to explore the gap between the solutions it provides and the ones generated by existing nonreactive approaches.