abstract: Reading partners’ actions correctly is essential for successful coordination, but interpretation does not always reflect reality. Attribution biases, such as self-serving and correspondence biases, lead people to misinterpret their partners’ actions and falsely assign blame after an unexpected event. These biases thus further influence people’s trust in their partners, including machine partners. The increasing capabilities and complexity of machines allow them to work physically with humans. However, their improvements may interfere with the accuracy for people to calibrate trust in machines and their capabilities, which requires an understanding of attribution biases’ effect on human-machine coordination. Specifically, the current thesis explores how the development of trust in a partner is influenced by attribution biases and people’s assignment of blame for a negative outcome. This study can also suggest how a machine partner should be designed to react to environmental disturbances and report the appropriate level of information about external conditions. / Dissertation/Thesis / Masters Thesis Human Systems Engineering 2019

Paul Dooley’s composition, Masks and Machines (2015), is a significant new work for wind ensemble and was the winner of the National Band Association’s William D. Revelli Memorial Band Composition Contest award and the American Bandmasters Association’s Sousa/ABA/Ostwald Composition Contest. Masks and Machines has received positive critical acclaim and numerous performances, including a performance at the 2015 Midwest Band and Orchestra Clinic in Chicago, Illinois by the North Texas Wind Ensemble under the direction of Eugene Corporon and a performance at the 2016 American Bandmasters Association Conference in Lexington, Kentucky by the United States Marine Corps Band under the direction of Jason K. Fettig. The purposes of this dissertation are 1) to place Masks and Machines in its historical perspective within the history of wind band compositions; 2) to provide an overview of the artistic styles that influenced the composer, such as Stravinsky’s Neoclassical works, Bauhaus Art, and Fortspinnung; 3) to elaborate on the musical traits and characteristics of Masks and Machines via formal analysis; and 4) to offer a guide to rehearsal and performance of the work. The Introduction discusses Masks and Machines in its historical context as a highly acclaimed wind ensemble composition within the canon of twentieth century wind band works. Chapter 1 includes a detailed biography of Paul Dooley. Chapter 2 discusses the visual art and musical influences on Paul Dooley and how these influences come to life in his wind band compositions. Chapter 3 is an analysis of Masks and Machines with thematic excerpts and discussions on form, instrumentation, orchestration, and compositional techniques. Chapter 4 provides a rehearsal and performance guide aimed to facilitate a successful performance of Masks and Machines. Chapter 5 includes a transcription of two interviews with the composer and focuses primarily on compositional influences, processes, and techniques regarding Masks and Machines and other wind band compositions by Dooley, such as Point Blank (2012), Meditation at Lagunitas (2014), and Mavericks (2016).

In this study, a Prediction Accuracy Based Hill Climbing Feature Selection Algorithm (AHCFS) is created and compared with an Error Rate Based Sequential Feature Selection Algorithm (ERFS) which is an existing Matlab algorithm. The goal of the study is to create a new piece of an algorithm that has potential to outperform the existing Matlab sequential feature selection algorithm in predicting the movement of S&P 500 (^GSPC) prices under certain circumstances. The two algorithms are tested based on historical data of ^GSPC, and Support Vector Machine (SVM) is employed by both as the classifier. A prediction without feature selection algorithm implemented is carried out and used as a baseline for comparison between the two algorithms. The prediction horizon set in this study for both algorithms varies from one to 60 days. The study results show that AHCFS reaches higher prediction accuracy than ERFS in the majority of the cases.

Extreme Learning Machine (ELM) is a training algorithm for Single-Layer Feed-forward Neural Network (SLFN). The difference in theory of ELM from other training algorithms is in the existence of explicitly-given solution due to the immutability of initialed weights. In practice, ELMs achieve performance similar to that of other state-of-the-art training techniques, while taking much less time to train a model. Experiments show that the speedup of training ELM is up to the 5 orders of magnitude comparing to standard Error Back-propagation algorithm. ELM is a recently discovered technique that has proved its efficiency in classic regression and classification tasks, including multi-class cases. In this thesis, extensions of ELMs for non-typical for Artificial Neural Networks (ANNs) problems are presented. The first extension, described in the third chapter, allows to use ELMs to get probabilistic outputs for multi-class classification problems. The standard way of solving this type of problems is based 'majority vote' of classifier's raw outputs. This approach can rise issues if the penalty for misclassification is different for different classes. In this case, having probability outputs would be more useful. In the scope of this extension, two methods are proposed. Additionally, an alternative way of interpreting probabilistic outputs is proposed. ELM method prove useful for non-linear dimensionality reduction and visualization, based on repetitive re-training and re-evaluation of model. The forth chapter introduces adaptations of ELM-based visualization for classification and regression tasks. A set of experiments has been conducted to prove that these adaptations provide better visualization results that can then be used for perform classification or regression on previously unseen samples. Shape registration of 3D models with non-isometric distortion is an open problem in 3D Computer Graphics and Computational Geometry. The fifth chapter discusses a novel approach for solving this problem by introducing a similarity metric for spectral descriptors. Practically, this approach has been implemented in two methods. The first one utilizes Siamese Neural Network to embed original spectral descriptors into a lower dimensional metric space, for which the Euclidean distance provides a good measure of similarity. The second method uses Extreme Learning Machines to learn similarity metric directly for original spectral descriptors. Over a set of experiments, the consistency of the proposed approach for solving deformable registration problem has been proven.

The objective of this study is to propose an efficient sampling-based RBDO using a new classification method to reduce the computational cost. In addition, accuracy improvement strategies for the Kriging method are proposed to reduce the number of expensive computer experiments. Current research effort involves: (1) developing a new classification method that is more efficient than conventional surrogate modeling methods while maintaining required accuracy level; (2) developing a sequential adaptive sampling method that inserts samples near the limit state function; (3) improving the efficiency of the RBDO process by using a fixed hyper-spherical local window with an efficient uniform sampling method and identification of active/violated constraints; and (4) improving the accuracy of the Kriging method by introducing several strategies. In the sampling-based RBDO, only accurate classification information is needed instead of accurate response surface. On the other hand, in general, surrogates are constructed using all available DoE samples instead of focusing on the limit state function. Therefore, the computational cost of surrogates can be relatively expensive; and the accuracy of the limit state (or decision) function can be sacrificed in return for reducing the error on unnecessary regions away from the limit state function. On the contrary, the support vector machine (SVM), which is a classification method, only uses support vectors, which are located near the limit state function, to focus on the decision function. Therefore, the SVM is very efficient and ideally applicable to sampling-based RBDO, if the accuracy of SVM is improved by inserting virtual samples near the limit state function. The proposed sequential sampling method inserts new samples near the limit state function so that the number of DoE samples is minimized. In many engineering problems, expensive computer simulations are used and thus the total computational cost needs to be reduced by using less number of DoE samples. Several efficiency strategies such as: (1) launching RBDO at a deterministic optimum design, (2) hyper-spherical local windows with an efficient uniform sampling method, (3) filtering of constraints, (4) sample reuse, (5) improved virtual sample generation, are used for the proposed sampling-based RBDO using virtual SVM. The number of computer experiments is also reduced by implementing accuracy improvement strategies for the Kriging method. Since the Kriging method is used for generating virtual samples and generating response surface of the cost function, the number of computer experiments can be reduced by introducing: (1) accurate correlation parameter estimation, (2) penalized maximum likelihood estimation (PMLE) for small sample size, (3) correlation model selection by MLE, and (4) mean structure selection by cross-validation (CV) error.

Learning a predictive model for a large scale real-world problem presents several challenges: the choice of a good feature set and a scalable machine learning algorithm with small generalization error. A support vector machine (SVM), based on statistical learning theory, obtains good generalization by restricting the capacity of its hypothesis space. A SVM outperforms classical learning algorithms on many benchmark data sets. Its excellent performance makes it the ideal choice for pattern recognition problems. However, training a SVM involves constrained quadratic programming, which leads to poor scalability. In this dissertation, we propose several methods to improve a SVM's scalability. The evaluation is done mainly in the context of a plankton recognition problem. One approach is called active learning, which selectively asks a domain expert to label a subset of examples from a lot of unlabeled data. Active learning minimizes the number of labeled examples needed to build an accurate model and reduces the human effort in manually labeling the data. We propose a new active learning method "Breaking Ties" (BT) for multi-class SVMs. After developing a probability model for multiple class SVMs, "BT" selectively labels examples for which the difference in probabilities between the predicted most likely class and second most likely class is smallest. This simple strategy required several times less labeled plankton images to reach a given recognition accuracy when compared to random sampling in our plankton recognition system. To speed up a SVM's training and prediction, we show how to apply bit reduction to compress the examples into several bins. Weights are assigned to different bins based on the number of examples in the bin. Treating each bin as a weighted example, a SVM builds a model using the reduced-set of weighted examples.

O presente estudo tem como objetivo fornecer subsídios para uma proposta de ensino- aprendizagem de conteúdos relativos às Transformações Geométricas planas com a utiliza- ção de software de Geometria Dinâmica, associado a Máquinas Matemáticas, visando como público-alvo alunos dos anos finais do Ensino Fundamental. Partindo de questões que envolvem tanto o ensino quanto a aprendizagem, ou mais especificamente, a apropriação do conceito de transformação geométrica por parte dos alunos, busca-se investigar em que medida o uso de um ambiente de Geometria Dinâmica associado ao uso de artefatos do tipo sistemas articulados pode contribuir para o desenvolvimento intelectual dos alunos nesse tema. O trabalho compreende um estudo bibliográfico sobre o tema, abordando as transformações geométricas como objetos matemáticos e objetos de ensino, culminando com a elaboração de uma proposta didática para a abordagem desse assunto na Educação Básica, com o uso dos recursos mencionados. / This study aims to provide support for a proposed teaching-learning content related to Plane Geometric Transformations with the use of Dynamic Geometry software, associated with Mathematical Machines, aiming target audience students of the final years of elementary school. Starting from issues involving both teaching and learning, or more specifically, the appropriation of the concept of geometric transformation by the students, we try to investigate to what extent the use of a Dynamic Geometry environment associated with the use of type linkages artifacts can contribute to the intellectual development of students in this subject. The work includes a literature study on the subject, addressing the geometric transformations as mathematical objects and learning objects, culminating in the development of a didactic proposal to address this matter in basic education, using the resources mentioned.

Sustainable energy solutions are aimed to reduce the consumption of fossil fuels by using renewable energy sources and energy efficiency techniques. This thesis presents two new sustainable energy applications in the field of electrical machines.Polyphase induction motors dominate the energy usage spectrum for industrial and commercial applications. The conventional winding structure used in both synchronous and induction machines has a basic unit of the winding with a 60 degree phase belt and a three phase connection either in star or delta. A new winding structure using an innovative Star-Delta Series Connection (SDSC) which has a high winding coefficient and low harmonic content is presented in this thesis. The principle of the SDSC winding is described. The Electro-Magnetic Belt and Electro-Magnetic Space diagram are two important means to be used for optimization of the new winding. Experimental results from two prototypes confirm the theoretical analysis. The efficiency of the new machine at rated load increased by about 3.8% as compared to the standard machine with a conventional winding structure.Wind energy is one of the most attractive renewable energy options. Wind turbines are designed to couple either synchronous or asynchronous generators with various forms of direct or indirect connection with grid or diesel generators. Permanent magnet (PM) generators using high energy Neodymium- Iron-Boron magnets offer advantages such as direct coupling without gear box, absence of excitation winding and slip rings, light weight and smaller size. This thesis presents the design and development of an outer-rotor PM generator suitable for wind energy conversion. The initial electromagnetic design followed by a Finite Element Analysis is presented in detail. A 20 kW prototype machine was built and extensively tested. It was found that the machine could maintain an ++ / efficiency of about 85% for a wide operating range. Equivalent circuit models were developed. The results of the Finite Element analysis matches closely with the experimental and the designed values.

Le développement des machines suit une logique liée aux performances, ce qui entraîne des niveaux de sollicitations toujours plus proches des limites matériaux. Cette tendance a pour effet de favoriser l'apparition de comportements vibratoires non-linéaires. Cette thèse a donc pour objectif d'étudier la dynamique non-linéaire des structures et tout particulièrement celle correspondant à des moteurs d'avion multi-rotor. Deux axes ont été développés en parallèle, le premier qui correspond aux aspects théoriques et le deuxième de nature expérimentale. En ce qui concerne les développements théoriques, nous nous sommes intéressés au cas des non-linéarités fortes conjuguées avec une richesse multi-fréquentielle des excitations. Dans ce cadre, nous avons étudié des modèles de roulements adaptés à la simulation en dynamique d'ensemble qui inclut les caractéristiques propres du contact ainsi que la présence des jeux. A l'aide de ces modèles, nous avons pu simuler les réponses vibratoires d'un système bi-rotor non-linéaire, notamment en contra-rotation. Afin de réaliser une étude pertinente de ce systËme nous avons développé un code qui généralise l'estimation des réponses au cas quasi-périodique et qui permet aussi de statuer sur la stabilité de ces régimes. Aussi, nous avons mis au point un banc d'essai bi-rotor représentatif du comportement dynamique d'un moteur réel. Après une phase de conception et de mise en \oe uvre du banc, différents types de tests de caractÈrisation ont été réalisés. Nous avons tout d'abord analysé les réponses vibratoires en co- et contra-rotation. Puis dans une deuxième phase, une configuration surcritique de l'arbre Basse Pression a été testée. L'ensemble de ces essais a permis de mettre en valeur les spécificités dynamiques des moteurs bi-rotor tant du point vue des régimes vibratoires (régimes périodiques, quasi-périodiques, modes directs et rétrogrades) que des effets non-linéaires (distorsion des pics, présence de surharmoniques d'ordre deux).

Ce travail concerne l'étude des machines asynchrones à cage dans les applications de traction automobile, en particulier pour les véhicules hybrides. Nous avons développé des modèles et des méthodes utiles pour analyser et dimensionner de telles machines électriques. Nous avons tout d'abord mis au point un modèle électromagnétique non linéaire de la machine, déterminé à partir d'un nombre restreint de calculs "éléments finis". Ce modèle a ensuite été adapté pour réaliser des études de sensibilités sur quelques dimensions géométriques importantes. Il permet d'adapter rapidement une machine à un nouveau cahier des charges. Nous avons finalement étendu cette méthode de modélisation pour prendre en compte un plus grand nombre de paramètres géométriques. Ce dernier modèle paramétrique a été utilisé pour l'optimisation sous contrainte des dimensions d'une machine. Pour cela, nous avons proposé une nouvelle méthode "d'optimisation à modèle recalé" qui concilie précision, rapidité et simplicité. Cette démarche a été appliquée au cas concret d'un dimensionnement de machine asynchrone avec un cahier des charges typique d'un véhicule hybride.