Enhancing Multimodal Silent Speech Interfaces With Feature Selection

In research on Silent Speech Interfaces (SSI), different sources of information (modalities) have been combined, aiming at obtaining better performance than the individual modalities. However, when combining these modalities, the dimensionality of the feature space rapidly increases, yielding the well-known "curse of dimensionality". As a consequence, in order to extract useful information from this data, one has to resort to feature selection (FS) techniques to lower the dimensionality of the learning space. In this paper, we assess the impact of FS techniques for silent speech data, in a dataset with 4 non-invasive and promising modalities, namely: video, depth, ultrasonic Doppler sensing, and surface electromyography. We consider two supervised (mutual information and Fisher's ratio) and two unsupervised (meanmedian and arithmetic mean geometric mean) FS filters. The evaluation was made by assessing the classification accuracy (word recognition error) of three well-known classifiers (knearest neighbors, support vector machines, and dynamic time warping). The key results of this study show that both unsupervised and supervised FS techniques improve on the classification accuracy on both individual and combined modalities. For instance, on the video component, we attain relative performance gains of 36.2% in error rates. FS is also useful as pre-processing for feature fusion. Copyright © 2014 ISCA.

Overcoming the objectification of western classical music with the concept of music as a process

Dissertation to obtain the degree of Master in Music - Artistic Interpretation

Drilling process of composite laminates – a tool based analysis

“Drilling of polymeric matrix composites structures”

Enhancing the selection of a model-based clustering with external categorical variables

In cluster analysis, it can be useful to interpret the partition built from the data in the light of external categorical variables which are not directly involved to cluster the data. An approach is proposed in the model-based clustering context to select a number of clusters which both fits the data well and takes advantage of the potential illustrative ability of the external variables. This approach makes use of the integrated joint likelihood of the data and the partitions at hand, namely the model-based partition and the partitions associated to the external variables. It is noteworthy that each mixture model is fitted by the maximum likelihood methodology to the data, excluding the external variables which are used to select a relevant mixture model only. Numerical experiments illustrate the promising behaviour of the derived criterion. © 2014 Springer-Verlag Berlin Heidelberg.

Characterization of the volatile fraction emitted by phloems of four pinus species by solid-phase microextraction and gas chromatography–mass spectrometry

Pine forests constitute some of the most important renewable resources supplying timber, paper and chemical industries, among other functions. Characterization of the volatiles emitted by different Pinus species has proven to be an important tool to decode the process of host tree selection by herbivore insects, some of which cause serious economic damage to pines. Variations in the relative composition of the bouquet of semiochemicals are responsible for the outcome of different biological processes, such as mate finding, egg-laying site recognition and host selection. The volatiles present in phloem samples of four pine species, P. halepensis, P. sylvestris, P. pinaster and P. pinea, were identified and characterized with the aim of finding possible host-plant attractants for native pests, such as the bark beetle Tomicus piniperda. The volatile compounds emitted by phloem samples of pines were extracted by headspace solid-phase micro extraction, using a 2 cm 50/30 mm divinylbenzene/carboxen/polydimethylsiloxane table flex solid-phase microextraction fiber and its contents analyzed by high-resolution gas chromatography, using flame ionization and a non polar and chiral column phases. The components of the volatile fraction emitted by the phloem samples were identified by mass spectrometry using time-of-flight and quadrupole mass analyzers. The estimated relative composition was used to perform a discriminant analysis among pine species, by means of cluster and principal component analysis. It can be concluded that it is possible to discriminate pine species based on the monoterpenes emissions of phloem samples.

An unsupervised approach to feature discretization and selection

Many learning problems require handling high dimensional datasets with a relatively small number of instances. Learning algorithms are thus confronted with the curse of dimensionality, and need to address it in order to be effective. Examples of these types of data include the bag-of-words representation in text classification problems and gene expression data for tumor detection/classification. Usually, among the high number of features characterizing the instances, many may be irrelevant (or even detrimental) for the learning tasks. It is thus clear that there is a need for adequate techniques for feature representation, reduction, and selection, to improve both the classification accuracy and the memory requirements. In this paper, we propose combined unsupervised feature discretization and feature selection techniques, suitable for medium and high-dimensional datasets. The experimental results on several standard datasets, with both sparse and dense features, show the efficiency of the proposed techniques as well as improvements over previous related techniques.

Efficient feature selection filters for high-dimensional data

Feature selection is a central problem in machine learning and pattern recognition. On large datasets (in terms of dimension and/or number of instances), using search-based or wrapper techniques can be cornputationally prohibitive. Moreover, many filter methods based on relevance/redundancy assessment also take a prohibitively long time on high-dimensional. datasets. In this paper, we propose efficient unsupervised and supervised feature selection/ranking filters for high-dimensional datasets. These methods use low-complexity relevance and redundancy criteria, applicable to supervised, semi-supervised, and unsupervised learning, being able to act as pre-processors for computationally intensive methods to focus their attention on smaller subsets of promising features. The experimental results, with up to 10(5) features, show the time efficiency of our methods, with lower generalization error than state-of-the-art techniques, while being dramatically simpler and faster.

TRAILER – e-Portfolio for collecting and recognizing informal learning process and outcomes

Informal learning is becoming more and more important: Nowadays people learn more this way, through the Internet, than in schools or normal trainings. But they don’t get any certificateswhich attest this fact. So they can't show the employer or teacher etc. that they have learned something. TRAILER project aim is to solve this problem by developing a special tool for managing of all competences and skills acquired through informal learning experiences. Both from the perspective of the user and the perspective of an institution or a company. We’ll present the IT tool to show how people can make their informal learning outcomes visible. TRAILER helps users to gather all information about process and outcomes of their informal learning. Users can share this with friends, colleagues or their employees, teachers and so on. They can create an interactive e-portfolio which can be attached to their CV, cover letter or Knowledge Management system etc. After the presentation of the tool we will discuss possible areas and fields to use this tool. Also we would like to discuss all possible use of the tool by the participants and another needs in this area. Moreover we want to discuss other problems in informal learning process, ways to solve the problems and discuss other ideas of different IT tools which could help in informal learning process. During the discussion we’ll use an interactive respond system which can be used on mobile devices: it makes possible for participants to share their opinions individually before knowing another persons' opinion.

Industry energy optimization: A case study in a biodiesel production site

This paper presents a case study of heat exchanger network (HEN) retrofit with the objective to reduce the utilities consumption in a biodiesel production process. Pinch analysis studies allow determining the minimum duty utilities as well the maximum of heat recovery. The existence of heat exchangers for heat recovery already running in the process causes a serious restriction for the implementation of grassroot HEN design based on pinch studies. Maintaining the existing HEN, a set of alternatives with additional heat exchangers was created and analysed using some industrial advice and selection criteria. The final proposed solution allows to increase the actual 18 % of recovery heat of the all heating needs of the process to 23 %, with an estimated annual saving in hot utility of 35 k(sic)/y.

Risk management of new product development process

Dissertation submitted to Faculdade de Ciências e Tecnologia of Universidade Nova de Lisboa for the achievement of Integrated Master´s degree in Industrial Management Engineering

A recursive process related to a partisan variation of Wythoff

Wythoff Queens is a classical combinatorial game related to very interesting mathematical results. An amazing one is the fact that the P-positions are given by (⌊├ φn⌋┤┤,├ ├ ⌊φ┤^2 n⌋) and (⌊├ φ^2 n⌋┤┤,├ ├ ⌊φ┤n⌋) where φ=(1+√5)/2. In this paper, we analyze a different version where one player (Left) plays with a chess bishop and the other (Right) plays with a chess knight. The new game (call it Chessfights) lacks a Beatty sequence structure in the P-positions as in Wythoff Queens. However, it is possible to formulate and prove some general results of a general recursive law which is a particular case of a Partizan Subtraction game.

Drilling damage in composite material

The characteristics of carbon fibre reinforced laminates have widened their use from aerospace to domestic appliances, and new possibilities for their usage emerge almost daily. In many of the possible applications, the laminates need to be drilled for assembly purposes. It is known that a drilling process that reduces the drill thrust force can decrease the risk of delamination. In this work, damage assessment methods based on data extracted from radiographic images are compared and correlated with mechanical test results—bearing test and delamination onset test—and analytical models. The results demonstrate the importance of an adequate selection of drilling tools and machining parameters to extend the life cycle of these laminates as a consequence of enhanced reliability.

A green integrated process for biodiesel production

Dissertation presented at Faculdade de Ciências e Tecnologia from Universidade Nova de Lisboa to obtain the degree of Master in Chemical and Biochemical Engineering

Land use/ land cover change and its impact on soil erosion process in Begnas Tal Rupa Tal watershed using geospatial tools, Kaski district, Nepal

Dissertation submitted in partial fulfilment of the requirements for the Degree of Master of Science in Geospatial Technologies

Estratégia de internacionalização da Grande Porto: seleção, modos de entrada e marketing-mix na abordagem dos mercados internacionais

Dissertação apresentada ao Instituto Politécnico do Porto para obtenção do Grau de Mestre em Gestão das Organizações, Ramo de Gestão de Empresas Orientada pelo Professor Doutor José Freitas Santos