Probabilistic topic models for sequence data

Probabilistic topic models are widely used in different contexts to uncover the hidden structure in large text corpora. One of the main (and perhaps strong) assumption of these models is that generative process follows a bag-of-words assumption, i.e. each token is independent from the previous one. We extend the popular Latent Dirichlet Allocation model by exploiting three different conditional Markovian assumptions: (i) the token generation depends on the current topic and on the previous token; (ii) the topic associated with each observation depends on topic associated with the previous one; (iii) the token generation depends on the current and previous topic. For each of these modeling assumptions we present a Gibbs Sampling procedure for parameter estimation. Experimental evaluation over real-word data shows the performance advantages, in terms of recall and precision, of the sequence-modeling approaches.     

Optimal radio resources and transmission formats assignment in OFDMA systems

The problem of assigning radio resources and transmission formats to users in the downlink of an OFDMA network is addressed. In particular, a single cell environment with a realistic interference model and a margin adaptive approach is considered, i.e., the aim is of minimizing the total transmission power while maintaining a certain given rate for each user. The computational complexity issues of the problem are discussed, and an approximation analysis is presented. Heuristic approaches, based on rounding techniques and graph models, are also proposed. Computational experiences show that, in a comparison with a commercial state-of-the-art optimization solver and with alternatives from the literature, the proposed algorithms are effective in terms of solution quality and CPU times.     

Biomechanics of fibrous proteins of the extracellular matrix studied by Brillouin scattering

Brillouin light scattering (BLS) spectroscopy is a technique that is able to detect thermally excited phonons within a material. The speed of propagation of these phonons can be determined from the magnitude of the Brillouin frequency shift between incident and scattered light, thereby providing a measure of the mechanical properties of the material in the gigahertz range. The mechanical properties of the extracellular matrices of biological tissues and their constituent biopolymers are important for normal tissue function and disturbances in these properties are widely implicated in disease. BLS offers the prospect of measuring mechanical properties on a microscopic scale in living tissues, thereby providing insights into structure–function relationships under normal and pathological conditions. In this study, we investigated BLS in collagen and elastin—the fibrous proteins of the extracellular matrix (ECM). Measurements were made on type I collagen in rat tail tendon, type II collagen in articular cartilage and nuchal ligament elastin. The dependence of the BLS spectrum on fibre orientation was investigated in a backscattering geometry using a reflective substrate. Two peaks, a bulk mode arising from phonon propagation along a quasi-radial direction to the fibre axis and a mode parallel to the surface, depending on sample orientation relative to the fibre axis, could be distinguished. The latter peak was fitted to a model of wave propagation through a hexagonally symmetric elastic solid, and the five components of the elasticity tensor were combined to give axial and transverse Young''s, shear and bulk moduli of the fibres. These were 10.2, 8.3, 3.2 and 10.9 GPa, and 6.1, 5.3, 1.9 and 8 GPa for dehydrated type I collagen and elastin, respectively. The former values are close to those previously reported. A microfocused BLS approach was also applied providing selection of single fibres. The moduli of collagen and elastin are much higher than those measured at lower frequency using macroscopic strains, and the difference between them is much less. We therefore believe, like previous investigators, that molecular-scale viscoelastic effects are responsible for the frequency dependence of the fibre biomechanics. Combining BLS with larger-scale mechanical testing methods therefore should, in the future, provide a means of following the evolution of mechanical properties in the formation of the complex structures found in the ECM.     

Periodic material structures tested by the noise spectroscopy method

The authors discuss the testing of periodic materials and structures via noise spectroscopy as an alternative to frequency sweep spectral analyzers, and the presented procedure employs the various advantages offered by a wide spectrum signal a Generally, noise spectroscopy tests are carried out to verify the behaviour of the response of periodic structures, and the related objective consists in recording the properties of microscopic structures in natural and artificial materials. The aim is to propose and verify a high-precision metrological method utilizable for the investigation of structures and materials in the frequency range between 100 MHz and 10 GHz; this paper therefore characterizes the design of a suitable measuring technique based on noise spectroscopy. In this context, the applied equipment is also shown to complement the underlying theoretical analysis.     

Design of integrated and autonomous conductivity–temperature–depth (CTD) sensors

Impedance characterization of interfaces is a basic technique for a large class of chemical and biological sensors. This technique is often used to model interfaces between ion-based and electron-based conductive materials by means of electric variables such as voltage, current and charge. Conductivity–temperature–depth (CTD) sensors are sophisticated devices used in the environmental monitoring field to understand the effects of climate changes on oceans and on marine organisms. They usually require impedance sensing as readout technique. High-accuracy CTD sensors are present on the market but they are bulky and power hungry. However, the downscale of modern CMOS technology allows shrinking very complex bioelectronic interfaces into millimeter square size systems, thus opening a large ground of applications. This paper will describe an IC architecture and the related design approach to implement an electrochemical impedance spectroscopy (EIS) technique for CTD sensing and will propose a general approach for sensing complex impedance with low power consumption and high precision. The presented system is designed to achieve 15-bit resolution and power consumption to ensure lifetime up to 1 year using button-size batteries in ocean environment.     

A Posteriori Analysis of a Positive Streamwise Invariant Discretization of a Convection-Diffusion Equation

We consider the finite element discretization of a convection-diffusion equation, where the convection term is handled via a fluctuation splitting algorithm. We prove a posteriori error estimates which allow us to perform mesh adaptivity in order to optimize the discretization of these equations. Numerical results confirm the interest of such an approach.     

Tapping into past design experiences: knowledge sharing and creation during novice–expert design consultations

Designing is a knowledge-intensive activity. For novice design engineers, an important means of acquiring knowledge is to consult experienced colleagues. We observed novice?Cexpert consultations as part of three engineering projects in a large aerospace company. Seven meetings were analysed in detail regarding the design activity, the content, and the form of interaction. Although the meetings were initiated for the purpose of information seeking, this process amounted to only 8% of the time compared to knowledge creation between novices and experts (47% of meeting time), and contextual information sharing (45% of meeting time). Both experts and novices were found to contribute equally and interactively to the discussion and analysis of solutions. The analysis showed how the processes alternated in the meetings. We identified tentative patterns on how these consultation processes change over the course of the design process phases. The micro-level analysis of the design activities and form of interaction provided a deeper understanding of how the consultation processes are discursively produced by the experts and novices. Finally, implications for design engineering practitioners are derived and suggestions for further research are provided.     

A structured overview on the use of computational simulators for teaching statistical methods

Modern technology is affecting what is taught and how statistics is taught. Computer intensive methods like bootstrapping, generalized regression, coordinate exchange algorithms and computer experiment emulators are now part of the statistical toolkit. In parallel with these new methods and tools, a similar evolution appears in statistical education. Simulators offer opportunities for hands-on experience in the classroom, bridging the gap between theory and practice. In this article, we map a wide range of simulation based tools used in teaching statistical methods, putting them in the context of a quality ladder representing various maturity levels of organizations and course level sophistication. The mapping is designed to help educators assess and integrate options related to the incorporation of simulators as teaching aids in an educational and pedagogical roadmap. The mapping is applied to specific simulation products but provides a general framework that can be expanded and used in a generic way.