Feature Selection for Speech Emotion Recognition in Spanish and Basque: On the Use of Machine Learning to Improve Human-Computer Interaction

Study of emotions in human-computer interaction is a growing research area. This paper shows an attempt to select the most significant features for emotion recognition in spoken Basque and Spanish Languages using different methods for feature selection. RekEmozio database was used as the experimental data set. Several Machine Learning paradigms were used for the emotion classification task. Experiments were executed in three phases, using different sets of features as classification variables in each phase. Moreover, feature subset selection was applied at each phase in order to seek for the most relevant feature subset. The three phases approach was selected to check the validity of the proposed approach. Achieved results show that an instance-based learning algorithm using feature subset selection techniques based on evolutionary algorithms is the best Machine Learning paradigm in automatic emotion recognition, with all different feature sets, obtaining a mean of 80,05% emotion recognition rate in Basque and a 74,82% in Spanish. In order to check the goodness of the proposed process, a greedy searching approach (FSS-Forward) has been applied and a comparison between them is provided. Based on achieved results, a set of most relevant non-speaker dependent features is proposed for both languages and new perspectives are suggested.

A feasible approach to optical-electrical hybrid data storage using phase change material

We present our experimental results supporting optical-electrical hybrid data storage by optical recording and electrical reading using Ge2Sb2Te5as recording medium. The sheet resistance of laser- irradiated Ge2Sb2Te5. lms exhibits an abrupt change of four orders of magnitude ( from 10 7 to 10 3./ sq) with increasing laser power, current- voltage curves of the amorphous area and the laser- crystallized dots, measured by a conductive atomic force microscope ( C- AFM), show that their resistivities are 2.725 and 3.375 x 10- 3., respectively, the surface current distribution in the. lms also shows high and low resistance states. All these results suggest that the laser- recorded bit can be read electrically by measuring the change of electrical resistivity, thus making optical electrical hybrid data storage possible.

Structural change of laser-irradiated Ge2Sb2Te5 films studied by electrical property measurement

Sheet resistance of laser-irradiated Ge2Sb2Te5 thin films prepared by magnetron sputtering was measured by the four-point probe method. With increasing laser power the sheet resistance undergoes an abrupt drop from 10(7) to 10(3) Omega/square at about 580 mW. The abrupt drop in resistance is due to the structural change from amorphous to crystalline state as revealed by X-ray diffraction (XRD) study of the samples around the abrupt change point. Crystallized dots were also formed in the amorphous Ge2Sb2Te5 films by focused short pulse laser-irradiated, the resistivities at the crystallized dots and the non-crystallized area are 3.375 x 10(-3) and 2.725 Omega m, sheet resistance is 3.37 x 10(4) and 2.725 x 10(7) Omega/square respectively, deduced from the I-V Curves that is obtained by conductive atomic force microscope (C-AFM). (C) 2008 Elsevier B.V. All rights reserved.

Instalaciones energéticas en edificios. viabilidad de las pilas de combustible en edificios de viviendas.

[ES]El siguiente proyecto analiza la viabilidad energética y económica que supone la instalación de una tecnología de pilas de combustible como cogeneración en una vivienda unifamiliar en Madrid. Al mismo tiempo, se compara dicha instalación con otra más desarrollada como es la combinación entre una caldera de condensación y la red eléctrica.

Gaussian processes for machine learning (GPML) toolbox

The GPML toolbox provides a wide range of functionality for Gaussian process (GP) inference and prediction. GPs are specified by mean and covariance functions; we offer a library of simple mean and covariance functions and mechanisms to compose more complex ones. Several likelihood functions are supported including Gaussian and heavy-tailed for regression as well as others suitable for classification. Finally, a range of inference methods is provided, including exact and variational inference, Expectation Propagation, and Laplace’s method dealing with non-Gaussian likelihoods and FITC for dealing with large regression tasks.

Structural, optical and electrical properties of ZnO films grown on c-plane sapphire and (100)gamma-LiAlO2 by pulse laser deposition

ZnO thin films were grown on single-crystal gamma-LiAlO2 (LAO) and sapphire (0001) substrate by pulsed laser deposition (PLD). The structural, optical and electrical properties of ZnO films were investigated. The results show that LAO is more suitable for fabricating ZnO films than sapphire substrate and the highest-quality ZnO film was attained on LAO at the substrate temperature of 550 degrees C. However, when the substrate temperature rises to 700 degrees C, lithium would diffuse from the substrate (LAO) into ZnO film which makes ZnO film on LAO becomes polycrystalline without preferred orientation, the stress in ZnO film increases dominantly and the resistivity of the film decreases exponentially. (c) 2005 Elsevier B.V. All rights reserved.

Directional algebraic reconstruction technique for electrical impedance tomography

We assume that the resistance matrix can be found in electrical impedance tomography from the assumption of linear dependence between the voltages and the currents and with the help of the resistance matrix and the transfer impedance between the electrodes, a directional algebraic reconstruction technique is proposed. The goal is to reconstruct the resistivity distribution by weighting the matrices that are obtained by calculating the orthogonal distance of the underlying mesh elements from the neighbouring port resistivity lines. These weighting matrices, which only depend on the topology of the underlying mesh, can be calculated offline and result in a computationally efficient online procedure with a reasonable image reconstruction performance. Simulation results are provided to validate this approach.