Performing Feature Selection with ACO

R. Jensen, 'Performing Feature Selection with ACO. Swarm Intelligence and Data Mining,' A. Abraham, C. Grosan and V. Ramos (eds.), Studies in Computational Intelligence, vol. 34, pp. 45-73. 2006.

Rough set based feature selection : A review

Feature selection aims to determine a minimal feature subset from a problem domain while retaining a suitably high accuracy in representing the original features. Rough set theory (RST) has been used as such a tool with much success. RST enables the discovery of data dependencies and the reduction of the number of attributes contained in a dataset using the data alone, requiring no additional information. This chapter describes the fundamental ideas behind RST-based approaches and reviews related feature selection methods that build on these ideas. Extensions to the traditional rough set approach are discussed, including recent selection methods based on tolerance rough sets, variable precision rough sets and fuzzy-rough sets. Alternative search mechanisms are also highly important in rough set feature selection. The chapter includes the latest developments in this area, including RST strategies based on hill-climbing, genetic algorithms and ant colony optimization.

Approximation-based feature selection and application for algae population estimation

Q. Shen and R. Jensen, 'Approximation-based feature selection and application for algae population estimation,' Applied Intelligence, vol. 28, no. 2, pp. 167-181, 2008. Sponsorship: EPSRC RONO: EP/E058388/1

Inspection of teats by colour image analysis for automatic milking systems

C.R. Bull, N.J.B. McFarlane, R. Zwiggelaar, C.J. Allen and T.T. Mottram, 'Inspection of teats by colour image analysis for automatic milking systems', Computers and Electronics in Agriculture 15 (1), 15-26 (1996)

ROSSI: Emergence of communication in Robots through Sensorimotor and Social Interaction

ROSSI: Emergence of communication in Robots through Sensorimotor and Social Interaction, T. Ziemke, A. Borghi, F. Anelli, C. Gianelli, F. Binkovski, G. Buccino, V. Gallese, M. Huelse, M. Lee, R. Nicoletti, D. Parisi, L. Riggio, A. Tessari, E. Sahin, International Conference on Cognitive Systems (CogSys 2008), University of Karlsruhe, Karlsruhe, Germany, 2008 Sponsorship: EU-FP7

Data Reduction with Rough Sets

R. Jensen, Q. Shen, Data Reduction with Rough Sets, In: Encyclopedia of Data Warehousing and Mining - 2nd Edition, Vol. II, 2008.

Age- and activity-related differences in the abundance of Myosin essential and regulatory light chains in human muscle

Traditional methods for phenotyping skeletal muscle (e.g., immunohistochemistry) are labor-intensive and ill-suited to multixplex analysis, i.e., assays must be performed in a series. Addressing these concerns represents a largely unmet research need but more comprehensive parallel analysis of myofibrillar proteins could advance knowledge regarding age- and activity-dependent changes in human muscle. We report a label-free, semi-automated and time efficient LC-MS proteomic workflow for phenotyping the myofibrillar proteome. Application of this workflow in old and young as well as trained and untrained human skeletal muscle yielded several novel observations that were subsequently verified by multiple reaction monitoring (MRM).We report novel data demonstrating that human ageing is associated with lesser myosin light chain 1 content and greater myosin light chain 3 content, consistent with an age-related reduction in type II muscle fibers. We also disambiguate conflicting data regarding myosin regulatory light chain, revealing that age-related changes in this protein more closely reflect physical activity status than ageing per se. This finding reinforces the need to control for physical activity levels when investigating the natural process of ageing. Taken together, our data confirm and extend knowledge regarding age- and activity-related phenotypes. In addition, the MRM transitions described here provide a methodological platform that can be fine-tuned to suite multiple research needs and thus advance myofibrillar phenotyping.

A survey of fingerprint classification Part I: taxonomies on feature extraction methods and learning models

This paper reviews the fingerprint classification literature looking at the problem from a double perspective. We first deal with feature extraction methods, including the different models considered for singular point detection and for orientation map extraction. Then, we focus on the different learning models considered to build the classifiers used to label new fingerprints. Taxonomies and classifications for the feature extraction, singular point detection, orientation extraction and learning methods are presented. A critical view of the existing literature have led us to present a discussion on the existing methods and their drawbacks such as difficulty in their reimplementation, lack of details or major differences in their evaluations procedures. On this account, an experimental analysis of the most relevant methods is carried out in the second part of this paper, and a new method based on their combination is presented.

Validation of a parabolic trough solar receiver model based on physical formulation of the absorber total emissivity from its spectral emissivity

Concentrating solar power is an important way of providing renewable energy. Model simulation approaches play a fundamental role in the development of this technology and, for this, an accurately validation of the models is crucial. This work presents the validation of the heat loss model of the absorber tube of a parabolic trough plant by comparing the model heat loss estimates with real measurements in a specialized testing laboratory. The study focuses on the implementation in the model of a physical-meaningful and widely valid formulation of the absorber total emissivity depending on the surface’s temperature. For this purpose, the spectral emissivity of several absorber’s samples are measured and, with these data, the absorber total emissivity curve is obtained according to Planck function. This physical-meaningful formulation is used as input parameter in the heat loss model and a successful validation of the model is performed. Since measuring the spectral emissivity of the absorber surface may be complex and it is sample-destructive, a new methodology for the absorber’s emissivity characterization is proposed. This methodology provides an estimation of the absorber total emissivity, retaining its physical meaning and widely valid formulation according to Planck function with no need for direct spectral measurements. This proposed method is also successfully validated and the results are shown in the present paper.

Análise das atitudes e comportamentos dos jovens adultos face aos iogurtes

Dissertação de Mestrado apresentada à Universidade Fernando Pessoa como parte dos requisitos para obtenção do grau de Mestre em Comunicação, especialização em Marketing e Publicidade.

A atitude dos consumidores face aos produtos alimentares de marca do distribuidor

Dissertação de Mestrado apresentada à Universidade Fernando Pessoa como parte dos requisitos para obtenção do grau de Mestre em Ciências da Comunicação, especialização em Marketing e Publicidade.

Um estudo sobre a personalidade da marca do distribuidor, no contexto da comunicação de produtos alimentares

Dissertação de Mestrado apresentada à Universidade Fernando Pessoa como parte dos requisitos para obtenção do grau de Mestre em Relações Públicas.

Gestão do cooperativismo médico no Brasil

Tese apresentada à Universidade Fernando Pessoa como parte dos requisitos para obtenção do grau de Doutor em Ciências Empresariais, especialidade em Gestão

Ergodicity and Mixing Rate of One-Dimensional Cellular Automata

One-and two-dimensional cellular automata which are known to be fault-tolerant are very complex. On the other hand, only very simple cellular automata have actually been proven to lack fault-tolerance, i.e., to be mixing. The latter either have large noise probability ε or belong to the small family of two-state nearest-neighbor monotonic rules which includes local majority voting. For a certain simple automaton L called the soldiers rule, this problem has intrigued researchers for the last two decades since L is clearly more robust than local voting: in the absence of noise, L eliminates any finite island of perturbation from an initial configuration of all 0's or all 1's. The same holds for a 4-state monotonic variant of L, K, called two-line voting. We will prove that the probabilistic cellular automata Kε and Lε asymptotically lose all information about their initial state when subject to small, strongly biased noise. The mixing property trivially implies that the systems are ergodic. The finite-time information-retaining quality of a mixing system can be represented by its relaxation time Relax(⋅), which measures the time before the onset of significant information loss. This is known to grow as (1/ε)^c for noisy local voting. The impressive error-correction ability of L has prompted some researchers to conjecture that Relax(Lε) = 2^(c/ε). We prove the tight bound 2^(c1log^21/ε) ＜ Relax(Lε) ＜ 2^(c2log^21/ε) for a biased error model. The same holds for Kε. Moreover, the lower bound is independent of the bias assumption. The strong bias assumption makes it possible to apply sparsity/renormalization techniques, the main tools of our investigation, used earlier in the opposite context of proving fault-tolerance.

Automatic Detection of Relevant Head Gestures in American Sign Language Communication

An automated system for detection of head movements is described. The goal is to label relevant head gestures in video of American Sign Language (ASL) communication. In the system, a 3D head tracker recovers head rotation and translation parameters from monocular video. Relevant head gestures are then detected by analyzing the length and frequency of the motion signal's peaks and valleys. Each parameter is analyzed independently, due to the fact that a number of relevant head movements in ASL are associated with major changes around one rotational axis. No explicit training of the system is necessary. Currently, the system can detect "head shakes." In experimental evaluation, classification performance is compared against ground-truth labels obtained from ASL linguists. Initial results are promising, as the system matches the linguists' labels in a significant number of cases.