Learning joint representations for order and timing of perceptual-motor sequences: a dynamic neural field approach

Many of our everyday tasks require the control of the serial order and the timing of component actions. Using the dynamic neural field (DNF) framework, we address the learning of representations that support the performance of precisely time action sequences. In continuation of previous modeling work and robotics implementations, we ask specifically the question how feedback about executed actions might be used by the learning system to fine tune a joint memory representation of the ordinal and the temporal structure which has been initially acquired by observation. The perceptual memory is represented by a self-stabilized, multi-bump activity pattern of neurons encoding instances of a sensory event (e.g., color, position or pitch) which guides sequence learning. The strength of the population representation of each event is a function of elapsed time since sequence onset. We propose and test in simulations a simple learning rule that detects a mismatch between the expected and realized timing of events and adapts the activation strengths in order to compensate for the movement time needed to achieve the desired effect. The simulation results show that the effector-specific memory representation can be robustly recalled. We discuss the impact of the fast, activation-based learning that the DNF framework provides for robotics applications.

Off-line simulation inspires insight: a neurodynamics approach to efficient robot task learning

There is currently an increasing demand for robots able to acquire the sequential organization of tasks from social learning interactions with ordinary people. Interactive learning-by-demonstration and communication is a promising research topic in current robotics research. However, the efficient acquisition of generalized task representations that allow the robot to adapt to different users and contexts is a major challenge. In this paper, we present a dynamic neural field (DNF) model that is inspired by the hypothesis that the nervous system uses the off-line re-activation of initial memory traces to incrementally incorporate new information into structured knowledge. To achieve this, the model combines fast activation-based learning to robustly represent sequential information from single task demonstrations with slower, weight-based learning during internal simulations to establish longer-term associations between neural populations representing individual subtasks. The efficiency of the learning process is tested in an assembly paradigm in which the humanoid robot ARoS learns to construct a toy vehicle from its parts. User demonstrations with different serial orders together with the correction of initial prediction errors allow the robot to acquire generalized task knowledge about possible serial orders and the longer term dependencies between subgoals in very few social learning interactions. This success is shown in a joint action scenario in which ARoS uses the newly acquired assembly plan to construct the toy together with a human partner.

Evaluating balance, stability and gait symmetry of stroke patients using kinematic and dynamic gait lab data

Dissertação de mestrado integrado em Engenharia Biomédica

“Learning-by-producing” in the fashion value chain

Proceedings da AUTEX 2015, Bucareste, Roménia.

La formación b-learning en el máster de español segunda lengua y lengua extranjera de la Universidade do Minho: innovación pedagógica y desafíos educativos

En el actual marco de creciente innovación pedagógica, debido entre otros factores, a la irrupción de nuevas herramientas informáticas, la enseñanza a distancia (e-learning y/o b-learning) va ocupando cada vez más espacio en la oferta educativa de diversas instituciones. En esta dirección, en la Universidade do Minho, y concretamente en el Área de Estudos Espanhóis e Hispano-Americanos,[1] hemos dedicado considerables esfuerzos a la ampliación de nuestra oferta desde 2010: primero en la elaboración e implementación del Curso de Formación Especializada en Español Lengua Extranjera, modalidad b-learning (3 ediciones; 2010-2013), y, actualmente, con el Máster Universitario en Español Lengua Segunda / Lengua Extranjera (vid. www.melsle.ilch.uminho.pt), también b-learning. En las siguientes páginas, nos proponemos compartir una serie de experiencias y reflexiones que han ido surgiendo durante estos años acerca de la formación universitária de profesores de Español Lengua Extranjera, en general, con recurso a la modalidade b-learning; para ello, nos centraremos en los siguientes aspectos: (i) caracterización general y problematización de la enseñanza a distancia en la Universidade do Minho; (ii) descripción del Máster Universitario en Español Lengua Segunda / Lengua Extranjera, acerca del cual detallaremos algunas prácticas adoptadas, relacionadas com la enseñanza e-learning como, por ejemplo, (iii) la coordinación pedagógica o (iv) los enfoques metodológicos adoptados a partir de la experiencia de una Unidad Curricular concreta.

Metabolic profiling and classification of propolis samples from Southern Brazil: an NMR-based platform coupled with machine learning

The chemical composition of propolis is affected by environmental factors and harvest season, making it difficult to standardize its extracts for medicinal usage. By detecting a typical chemical profile associated with propolis from a specific production region or season, certain types of propolis may be used to obtain a specific pharmacological activity. In this study, propolis from three agroecological regions (plain, plateau, and highlands) from southern Brazil, collected over the four seasons of 2010, were investigated through a novel NMR-based metabolomics data analysis workflow. Chemometrics and machine learning algorithms (PLS-DA and RF), including methods to estimate variable importance in classification, were used in this study. The machine learning and feature selection methods permitted construction of models for propolis sample classification with high accuracy (>75%, reaching 90% in the best case), better discriminating samples regarding their collection seasons comparatively to the harvest regions. PLS-DA and RF allowed the identification of biomarkers for sample discrimination, expanding the set of discriminating features and adding relevant information for the identification of the class-determining metabolites. The NMR-based metabolomics analytical platform, coupled to bioinformatic tools, allowed characterization and classification of Brazilian propolis samples regarding the metabolite signature of important compounds, i.e., chemical fingerprint, harvest seasons, and production regions.

Investigation on the homogeneity and stability of aqueous nano cellulose suspensions prepared using pluronic F-127

This paper presents the results of experimental investigation on the aqueous dispersion behaviour of micro crystalline cellulose (MCC) prepared using Pluronic F-127. For this purpose, different concentrations (0.5-3.0 wt.%) of MCC were dispersed in water with the help of ultrasonication technique using various concentrations of Pluronic F-127. The homogeneity of the suspensions and agglomerations were characterized by optical and transmission electron microscopy and the concentration of well dispersed MCC was measured using UV-Vis spectroscopy. Also, the suspensions were subjected to high speed ultracentrifugation at 3000 rpm and observed visually for sedimentation and subsequently, concentration was calculated using UV-Vis, in order to assess the long term stability of the suspensions. Based on these experiments, optimum concentration of Pluronic to disperse different MCC concentrations has been suggested.

Hand gesture recognition system based in computer vision and machine learning: Applications on human-machine interaction

Tese de Doutoramento em Engenharia de Eletrónica e de Computadores

“Learning through Radio, Learning for Life!”: Notas sobre o desenvolvimento de uma rádio participativa online

(Excerto) Olhando para o percurso do RadioActive, há uma ideia que parece ser transversal a todo o projeto. Referimo-nos a um princípio que chamaríamos de “identificação” e que foi determinante – é determinante – nos processos de investigação participativa. Falamos da identificação dos investigadores com os princípios da investigação-ação, da identificação das intervenções com as particularidades de cada contexto. Da imprescindível e progressiva identificação dos participantes com o projeto. Na verdade, sem esta multifacetada identificação é impossível pensar em resultados sustentáveis e persistentes. Investigadores e demais participantes têm de sentir que o projeto é “seu”, que os objetivos são “seus”, embora o façam necessariamente a velocidades diferentes. A aprendizagem, neste âmbito, expande-se sempre de dentro para fora, emerge dos interesses do sujeito e não de uma estrutura pré-concebida e imposta pelos que chegam (Ravenscroft et al., 2011), neste caso, os investigadores. Uma das diferenças das pesquisas participativas em relação às tradicionais é, precisamente, a atuação coletiva e não solitária do investigador. Os pesquisadores fazem parte de um processo participatório em que estão envolvidos numa estrutura (Cammarota & Fine, 2008: 5). Paulo Freire é o autor primordial em todos os projetos e países onde a RA101 foi aplicada. As suas concepções em torno da investigação-ação participativa tentam apontar sempre para uma ação e também para uma reflexão sobre os processos.

In vitro digestion and stability assessment of b-lactoglobulin/riboflavin nanostructures

B-Lactoglobulin (b-Lg) is the major protein fraction of bovine whey serum and a primary gelling agent. b-Lg has a high nutritional value, is stable at low pH being highly resistant to proteolytic degradation in the stomach, besides, it has the ability of acting as an encapsulating agent. This study aims at assessing the ability of b-Lg nanostructures to associate a nutraceutical - i.e. riboflavin - and release it in a controlled manner throughout an in vitro gastrointestinal (GI) system. For this reason b-Lg nanostructures loaded with riboflavin were critically characterized in terms of their morphology (i.e. size, polydispersity, -potential and shape) by dynamic light scattering (DLS) and transmission electron microscopy (TEM), and efficiency to associate to riboflavin through spectrofluorimetry. Furthermore, these nanocomplexes were evaluated in an in vitro GI model, simulating the physiological conditions. Stable b-Lg nanostructures were obtained at pH 6, of spherical shape, characterized by particle size of 172±1 nm, low polydispersity (i.e. PDI of 0.06±0.02), -potential of -32±3 mV and association efficiency (AE) of 26±1 %. b-Lg nanostructures showed to be stable upon their passage throughout stomach (i.e. particle size, PDI and potential of 248±10 nm, 0.18±0.03 and 18±3 mV, respectively). Concerning their passage throughout the intestine, such nanostructures were mostly degraded in the duodenum. Regarding riboflavin, a release of about 11 % was observed after their passage through stomach, while 35 %, 38 % and 5 % were the released percentages of the total riboflavin associated observed after passage through duodenum, jejunum and ileum, respectively. Hence,b-Lg nanostructures showed to be suitable carriers for riboflavin until the intestine, where their degradation occurs. b-Lg also showed to be structurally stable, under food simulant conditions (yoghurt simulant, composed of 3 % acetic acid), over 14 days, with a protective effect upon riboflavin activity, releasing it in a 7 day period.

Engineering a commercial cold-adapted xylanase for improved low pH stability

Dissertação de mestrado em Genética Molecular

O desenvolvimento do pensamento abstrato através de game-based learning e do Scratch

Relatório de estágio de mestrado em Ensino de Informática

Laccase immobilization on bacterial nanocellulose membranes: antimicrobial, kinetic and stability properties

This work studied the physical immobilization of a commercial laccase on bacterial nanocellulose (BNC) aiming to identify the laccase antibacterial properties suitable for wound dressings. Physico-chemical analysis demonstrates that the BNC structure is manly formed by pure crystalline I cellulose. The pH optimum and activation energy of free laccase depends on the substrate employed corresponding to pH 6, 7, 3 and 57, 22, 48 kJ mol1 for 2,6-dimethylphenol (DMP), catechol and 2,2 -azino-bis-(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS), respectively. The Michaelis-Menten constant (Km) value for the immobilized laccase (0.77 mM) was found to be almost double of that of the free enzyme (0.42 mM). However, the specific activities of immobilized and free laccase are similar suggesting that the cage-like structure of BNC allows entrapped laccase to maintain some flexibility and favour substrate accessibility. The results clearly show the antimicrobial effect of laccase in Gram-positive (92%) and Gram-negative (26%) bacteria and cytotoxicity acceptable for wound dressing applications.

Educating for sociability and learning for citizenship: developing communicational skills and mediation in educational contexts

This text concerns a program about the Promotion of Social and Communicational Skills and Mediation (PSCSM) developed with children aged between 10 and 13 years in a non-formal educational institution. The program of intervention had, as its purpose, the promotion of social and communicational competencies and mediation, thus enabling the children involved to have a healthy and responsible sociability in the different contexts in which they find themselves: family, school, peer group, amongst others. It was developed over 13 sessions with objectives and activities intentionally planned with the view of promoting competencies of communication, co-operation, responsibility, a critical spirit, solidarity, autonomy, respect, integration, inclusion and the recognition of rights and duties. This work was carried out with an action-research methodology that resorted to various techniques and instruments to gather and record information. The results obtained showed the impact and benefits of the program and they also revealed the necessity of educational institutions investing in the promotion of an ethical literacy and the empowerment of the children and young people for healthy sociability and active citizenship.