Implications for the scheduling of theory versus application in learning

This paper presents the findings of an experiment which looked at the effects of performing applied tasks (action learning) prior to the completion of the theoretical learning of these tasks (explanation-based learning), and vice-versa. The applied tasks took the form of laboratories for the Object-Oriented Analysis and Design (OOAD) course, theoretical learning was via lectures.

Machine learning for transient discovery in Pan-STARRS1 difference imaging

Efficient identification and follow-up of astronomical transients is hindered by the need for humans to manually select promising candidates from data streams that contain many false positives. These artefacts arise in the difference images that are produced by most major ground-based time-domain surveys with large format CCD cameras. This dependence on humans to reject bogus detections is unsustainable for next generation all-sky surveys and significant effort is now being invested to solve the problem computationally. In this paper, we explore a simple machine learning approach to real-bogus classification by constructing a training set from the image data of similar to 32 000 real astrophysical transients and bogus detections from the Pan-STARRS1 Medium Deep Survey. We derive our feature representation from the pixel intensity values of a 20 x 20 pixel stamp around the centre of the candidates. This differs from previous work in that it works directly on the pixels rather than catalogued domain knowledge for feature design or selection. Three machine learning algorithms are trained (artificial neural networks, support vector machines and random forests) and their performances are tested on a held-out subset of 25 per cent of the training data. We find the best results from the random forest classifier and demonstrate that by accepting a false positive rate of 1 per cent, the classifier initially suggests a missed detection rate of around 10 per cent. However, we also find that a combination of bright star variability, nuclear transients and uncertainty in human labelling means that our best estimate of the missed detection rate is approximately 6 per cent.

Strange, But True? Object-oriented Programming Is Best Taught, And Learnt, While Sitting On The Floor.

There is a perception amongst some of those learning computer programming that the principles of object-oriented programming (where behaviour is often encapsulated across multiple class files) can be difficult to grasp, especially when taught through a traditional, didactic ‘talk-and-chalk’ method or in a lecture-based environment.
We propose a non-traditional teaching method, developed for a government funded teaching training project delivered by Queen’s University, we call it bigCode. In this scenario, learners are provided with many printed, poster-sized fragments of code (in this case either Java or C#). The learners sit on the floor in groups and assemble these fragments into the many classes which make-up an object-oriented program.
Early trials indicate that bigCode is an effective method for teaching object-orientation. The requirement to physically organise the code fragments imitates closely the thought processes of a good software developer when developing object-oriented code.
Furthermore, in addition to teaching the principles involved in object-orientation, bigCode is also an extremely useful technique for teaching learners the organisation and structure of individual classes in Java or C# (as well as the organisation of procedural code). The mechanics of organising fragments of code into complete, correct computer programs give the users first-hand practice of this important skill, and as a result they subsequently find it much easier to develop well-structured code on a computer.
Yet, open questions remain. Is bigCode successful only because we have unknowingly predominantly targeted kinesthetic learners? Is bigCode also an effective teaching approach for other forms of learners, such as visual learners? How scalable is bigCode: in its current form can it be used with large class sizes, or outside the classroom?

Experiential learning instruments : an examination of the concept of the instrumental approach in experiential learning

Experiential Learning Instruments (ELls) are employed to modify the leamer's apprehension and / or comprehension in experiential learning situations, thereby improving the efficiency and effectiveness of those modalities in the learning process. They involve the learner in reciprocally interactive and determining transactions with his/her environment. Experiential Learning Instruments are used to keep experiential learning a process rather than an object. Their use is aimed at the continual refinement of the learner's knowledge and skill. Learning happens as the leamer's awareness, directed by the use of Ells, comes to experience, monitor and then use experiential feedback from living situations in a way that facilitates knmvledge/skill acquisition, self-correction and refinement. The thesis examined the literature relevant to the establishing of a theoretical experiential learning framework within which ELls can be understood. This framework included the concept that some learnings have intrinsic value-knowledge of necessary information-while others have instrumental value-knowledge of how to learn. The Kolb Learning Cycle and Kolb's six characteristics of experiential learning were used in analyzing three ELls from different fields of learning-saxophone tone production, body building and interpersonal communications. The ELls were examined to determine their learning objectives and how they work using experiential learning situations. It was noted that ELls do not transmit information but assist the learner in attending to and comprehending aspects of personal experience. Their function is to telescope the experiential learning process.

Disambiguating the Effects of Social Instability Stress in Adolescence on Learning and Memory Tasks that Involve the Medial Prefrontal Cortex and Hippocampus

Exposure to chronic stress can alter the structure and function of brain regions involved in learning and memory, and these effects are typically long-lasting if the stress occurs during sensitive periods of development. Until recently, adolescence has received relatively little attention as a sensitive period of development, despite marked changes in behaviour, heightened reactivity to stressors, and cognitive and neural maturation. Therefore, the purpose of the present study was to investigate the long-term effects of chronic stress in adolescence on two spatial learning and memory tasks (Morris water maze and Spatial Object Location test) and on a working memory task (Delayed Alternation task). Male rats were randomly assigned to chronic social instability stress (SS; daily 1 hour isolation and subsequent change of cage partner between postnatal days 30 and 45) or to a no-stress control group (CTL). During acquisition learning in the Morris water maze task, SS rats demonstrated impaired long-term memory for the location of the hidden escape platform compared to CTL rats, although the impairment was only seen after the first day of training. Similarly, SS rats had impaired long-term memory in the Spatial Object Location test after a long delay (240 minutes), but not after shorter delays (15 or 60 minutes) compared to CTL rats. On the Delayed Alternation task, which assessed working memory across delays ranging from 5 to 90 seconds, no group differences were observed. These results are partially in line with previous research that revealed adult impairment on spatial learning and memory tasks after exposure to chronic social instability stress in adolescence. The observed deficits, however, appear to be limited to long-term memory as no group differences were observed during brief periods of retention.

Patterns and quality of object-oriented software systems

Lors de ces dix dernières années, le coût de la maintenance des systèmes orientés objets s'est accru jusqu' à compter pour plus de 70% du coût total des systèmes. Cette situation est due à plusieurs facteurs, parmi lesquels les plus importants sont: l'imprécision des spécifications des utilisateurs, l'environnement d'exécution changeant rapidement et la mauvaise qualité interne des systèmes. Parmi tous ces facteurs, le seul sur lequel nous ayons un réel contrôle est la qualité interne des systèmes. De nombreux modèles de qualité ont été proposés dans la littérature pour contribuer à contrôler la qualité. Cependant, la plupart de ces modèles utilisent des métriques de classes (nombre de méthodes d'une classe par exemple) ou des métriques de relations entre classes (couplage entre deux classes par exemple) pour mesurer les attributs internes des systèmes. Pourtant, la qualité des systèmes par objets ne dépend pas uniquement de la structure de leurs classes et que mesurent les métriques, mais aussi de la façon dont celles-ci sont organisées, c'est-à-dire de leur conception, qui se manifeste généralement à travers les patrons de conception et les anti-patrons. Dans cette thèse nous proposons la méthode DEQUALITE, qui permet de construire systématiquement des modèles de qualité prenant en compte non seulement les attributs internes des systèmes (grâce aux métriques), mais aussi leur conception (grâce aux patrons de conception et anti-patrons). Cette méthode utilise une approche par apprentissage basée sur les réseaux bayésiens et s'appuie sur les résultats d'une série d'expériences portant sur l'évaluation de l'impact des patrons de conception et des anti-patrons sur la qualité des systèmes. Ces expériences réalisées sur 9 grands systèmes libres orientés objet nous permettent de formuler les conclusions suivantes: • Contre l'intuition, les patrons de conception n'améliorent pas toujours la qualité des systèmes; les implantations très couplées de patrons de conception par exemple affectent la structure des classes et ont un impact négatif sur leur propension aux changements et aux fautes. • Les classes participantes dans des anti-atrons sont beaucoup plus susceptibles de changer et d'être impliquées dans des corrections de fautes que les autres classes d'un système. • Un pourcentage non négligeable de classes sont impliquées simultanément dans des patrons de conception et dans des anti-patrons. Les patrons de conception ont un effet positif en ce sens qu'ils atténuent les anti-patrons. Nous appliquons et validons notre méthode sur trois systèmes libres orientés objet afin de démontrer l'apport de la conception des systèmes dans l'évaluation de la qualité.

Deep learning of representations and its application to computer vision

L’objectif de cette thèse par articles est de présenter modestement quelques étapes du parcours qui mènera (on espère) à une solution générale du problème de l’intelligence artificielle. Cette thèse contient quatre articles qui présentent chacun une différente nouvelle méthode d’inférence perceptive en utilisant l’apprentissage machine et, plus particulièrement, les réseaux neuronaux profonds. Chacun de ces documents met en évidence l’utilité de sa méthode proposée dans le cadre d’une tâche de vision par ordinateur. Ces méthodes sont applicables dans un contexte plus général, et dans certains cas elles on tété appliquées ailleurs, mais ceci ne sera pas abordé dans le contexte de cette de thèse. Dans le premier article, nous présentons deux nouveaux algorithmes d’inférence variationelle pour le modèle génératif d’images appelé codage parcimonieux “spike- and-slab” (CPSS). Ces méthodes d’inférence plus rapides nous permettent d’utiliser des modèles CPSS de tailles beaucoup plus grandes qu’auparavant. Nous démontrons qu’elles sont meilleures pour extraire des détecteur de caractéristiques quand très peu d’exemples étiquetés sont disponibles pour l’entraînement. Partant d’un modèle CPSS, nous construisons ensuite une architecture profonde, la machine de Boltzmann profonde partiellement dirigée (MBP-PD). Ce modèle a été conçu de manière à simplifier d’entraînement des machines de Boltzmann profondes qui nécessitent normalement une phase de pré-entraînement glouton pour chaque couche. Ce problème est réglé dans une certaine mesure, mais le coût d’inférence dans le nouveau modèle est relativement trop élevé pour permettre de l’utiliser de manière pratique. Dans le deuxième article, nous revenons au problème d’entraînement joint de machines de Boltzmann profondes. Cette fois, au lieu de changer de famille de modèles, nous introduisons un nouveau critère d’entraînement qui donne naissance aux machines de Boltzmann profondes à multiples prédictions (MBP-MP). Les MBP-MP sont entraînables en une seule étape et ont un meilleur taux de succès en classification que les MBP classiques. Elles s’entraînent aussi avec des méthodes variationelles standard au lieu de nécessiter un classificateur discriminant pour obtenir un bon taux de succès en classification. Par contre, un des inconvénients de tels modèles est leur incapacité de générer deséchantillons, mais ceci n’est pas trop grave puisque la performance de classification des machines de Boltzmann profondes n’est plus une priorité étant donné les dernières avancées en apprentissage supervisé. Malgré cela, les MBP-MP demeurent intéressantes parce qu’elles sont capable d’accomplir certaines tâches que des modèles purement supervisés ne peuvent pas faire, telles que celle de classifier des données incomplètes ou encore celle de combler intelligemment l’information manquante dans ces données incomplètes. Le travail présenté dans cette thèse s’est déroulé au milieu d’une période de transformations importantes du domaine de l’apprentissage à réseaux neuronaux profonds qui a été déclenchée par la découverte de l’algorithme de “dropout” par Geoffrey Hinton. Dropout rend possible un entraînement purement supervisé d’architectures de propagation unidirectionnel sans être exposé au danger de sur- entraînement. Le troisième article présenté dans cette thèse introduit une nouvelle fonction d’activation spécialement con ̧cue pour aller avec l’algorithme de Dropout. Cette fonction d’activation, appelée maxout, permet l’utilisation de aggrégation multi-canal dans un contexte d’apprentissage purement supervisé. Nous démontrons comment plusieurs tâches de reconnaissance d’objets sont mieux accomplies par l’utilisation de maxout. Pour terminer, sont présentons un vrai cas d’utilisation dans l’industrie pour la transcription d’adresses de maisons à plusieurs chiffres. En combinant maxout avec une nouvelle sorte de couche de sortie pour des réseaux neuronaux de convolution, nous démontrons qu’il est possible d’atteindre un taux de succès comparable à celui des humains sur un ensemble de données coriace constitué de photos prises par les voitures de Google. Ce système a été déployé avec succès chez Google pour lire environ cent million d’adresses de maisons.

FPGA-based object detection using classification circuits

Dans l'apprentissage machine, la classification est le processus d’assigner une nouvelle observation à une certaine catégorie. Les classifieurs qui mettent en œuvre des algorithmes de classification ont été largement étudié au cours des dernières décennies. Les classifieurs traditionnels sont basés sur des algorithmes tels que le SVM et les réseaux de neurones, et sont généralement exécutés par des logiciels sur CPUs qui fait que le système souffre d’un manque de performance et d’une forte consommation d'énergie. Bien que les GPUs puissent être utilisés pour accélérer le calcul de certains classifieurs, leur grande consommation de puissance empêche la technologie d'être mise en œuvre sur des appareils portables tels que les systèmes embarqués. Pour rendre le système de classification plus léger, les classifieurs devraient être capable de fonctionner sur un système matériel plus compact au lieu d'un groupe de CPUs ou GPUs, et les classifieurs eux-mêmes devraient être optimisés pour ce matériel. Dans ce mémoire, nous explorons la mise en œuvre d'un classifieur novateur sur une plate-forme matérielle à base de FPGA. Le classifieur, conçu par Alain Tapp (Université de Montréal), est basé sur une grande quantité de tables de recherche qui forment des circuits arborescents qui effectuent les tâches de classification. Le FPGA semble être un élément fait sur mesure pour mettre en œuvre ce classifieur avec ses riches ressources de tables de recherche et l'architecture à parallélisme élevé. Notre travail montre que les FPGAs peuvent implémenter plusieurs classifieurs et faire les classification sur des images haute définition à une vitesse très élevée.

Object Recognition with Pictorial Structures

This thesis presents a statistical framework for object recognition. The framework is motivated by the pictorial structure models introduced by Fischler and Elschlager nearly 30 years ago. The basic idea is to model an object by a collection of parts arranged in a deformable configuration. The appearance of each part is modeled separately, and the deformable configuration is represented by spring-like connections between pairs of parts. These models allow for qualitative descriptions of visual appearance, and are suitable for generic recognition problems. The problem of detecting an object in an image and the problem of learning an object model using training examples are naturally formulated under a statistical approach. We present efficient algorithms to solve these problems in our framework. We demonstrate our techniques by training models to represent faces and human bodies. The models are then used to locate the corresponding objects in novel images.

Teaching an Old Robot New Tricks: Learning Novel Tasks via Interaction with People and Things

As AI has begun to reach out beyond its symbolic, objectivist roots into the embodied, experientialist realm, many projects are exploring different aspects of creating machines which interact with and respond to the world as humans do. Techniques for visual processing, object recognition, emotional response, gesture production and recognition, etc., are necessary components of a complete humanoid robot. However, most projects invariably concentrate on developing a few of these individual components, neglecting the issue of how all of these pieces would eventually fit together. The focus of the work in this dissertation is on creating a framework into which such specific competencies can be embedded, in a way that they can interact with each other and build layers of new functionality. To be of any practical value, such a framework must satisfy the real-world constraints of functioning in real-time with noisy sensors and actuators. The humanoid robot Cog provides an unapologetically adequate platform from which to take on such a challenge. This work makes three contributions to embodied AI. First, it offers a general-purpose architecture for developing behavior-based systems distributed over networks of PC's. Second, it provides a motor-control system that simulates several biological features which impact the development of motor behavior. Third, it develops a framework for a system which enables a robot to learn new behaviors via interacting with itself and the outside world. A few basic functional modules are built into this framework, enough to demonstrate the robot learning some very simple behaviors taught by a human trainer. A primary motivation for this project is the notion that it is practically impossible to build an "intelligent" machine unless it is designed partly to build itself. This work is a proof-of-concept of such an approach to integrating multiple perceptual and motor systems into a complete learning agent.

Example Based Learning for View-Based Human Face Detection

We present an example-based learning approach for locating vertical frontal views of human faces in complex scenes. The technique models the distribution of human face patterns by means of a few view-based "face'' and "non-face'' prototype clusters. At each image location, the local pattern is matched against the distribution-based model, and a trained classifier determines, based on the local difference measurements, whether or not a human face exists at the current image location. We provide an analysis that helps identify the critical components of our system.

View-Based Strategies for 3D Object Recognition

A persistent issue of debate in the area of 3D object recognition concerns the nature of the experientially acquired object models in the primate visual system. One prominent proposal in this regard has expounded the use of object centered models, such as representations of the objects' 3D structures in a coordinate frame independent of the viewing parameters [Marr and Nishihara, 1978]. In contrast to this is another proposal which suggests that the viewing parameters encountered during the learning phase might be inextricably linked to subsequent performance on a recognition task [Tarr and Pinker, 1989; Poggio and Edelman, 1990]. The 'object model', according to this idea, is simply a collection of the sample views encountered during training. Given that object centered recognition strategies have the attractive feature of leading to viewpoint independence, they have garnered much of the research effort in the field of computational vision. Furthermore, since human recognition performance seems remarkably robust in the face of imaging variations [Ellis et al., 1989], it has often been implicitly assumed that the visual system employs an object centered strategy. In the present study we examine this assumption more closely. Our experimental results with a class of novel 3D structures strongly suggest the use of a view-based strategy by the human visual system even when it has the opportunity of constructing and using object-centered models. In fact, for our chosen class of objects, the results seem to support a stronger claim: 3D object recognition is 2D view-based.

Observations on Cortical Mechanisms for Object Recognition andsLearning

This paper sketches a hypothetical cortical architecture for visual 3D object recognition based on a recent computational model. The view-centered scheme relies on modules for learning from examples, such as Hyperbf-like networks. Such models capture a class of explanations we call Memory-Based Models (MBM) that contains sparse population coding, memory-based recognition, and codebooks of prototypes. Unlike the sigmoidal units of some artificial neural networks, the units of MBMs are consistent with the description of cortical neurons. We describe how an example of MBM may be realized in terms of cortical circuitry and biophysical mechanisms, consistent with psychophysical and physiological data.

Object Detection in Images by Components

In this paper we present a component based person detection system that is capable of detecting frontal, rear and near side views of people, and partially occluded persons in cluttered scenes. The framework that is described here for people is easily applied to other objects as well. The motivation for developing a component based approach is two fold: first, to enhance the performance of person detection systems on frontal and rear views of people and second, to develop a framework that directly addresses the problem of detecting people who are partially occluded or whose body parts blend in with the background. The data classification is handled by several support vector machine classifiers arranged in two layers. This architecture is known as Adaptive Combination of Classifiers (ACC). The system performs very well and is capable of detecting people even when all components of a person are not found. The performance of the system is significantly better than a full body person detector designed along similar lines. This suggests that the improved performance is due to the components based approach and the ACC data classification structure.

Learning Three-Dimensional Shape Models for Sketch Recognition

Artifacts made by humans, such as items of furniture and houses, exhibit an enormous amount of variability in shape. In this paper, we concentrate on models of the shapes of objects that are made up of fixed collections of sub-parts whose dimensions and spatial arrangement exhibit variation. Our goals are: to learn these models from data and to use them for recognition. Our emphasis is on learning and recognition from three-dimensional data, to test the basic shape-modeling methodology. In this paper we also demonstrate how to use models learned in three dimensions for recognition of two-dimensional sketches of objects.