Studying Recommender Systems to Enhance Distributed Computing Schedulers

Distributed Computing frameworks belong to a class of programming models that allow developers to

launch workloads on large clusters of machines. Due to the dramatic increase in the volume of

data gathered by ubiquitous computing devices, data analytic workloads have become a common

case among distributed computing applications, making Data Science an entire field of

Computer Science. We argue that Data Scientist's concern lays in three main components: a dataset,

a sequence of operations they wish to apply on this dataset, and some constraint they may have

related to their work (performances, QoS, budget, etc). However, it is actually extremely

difficult, without domain expertise, to perform data science. One need to select the right amount

and type of resources, pick up a framework, and configure it. Also, users are often running their

application in shared environments, ruled by schedulers expecting them to specify precisely their resource

needs. Inherent to the distributed and concurrent nature of the cited frameworks, monitoring and

profiling are hard, high dimensional problems that block users from making the right

configuration choices and determining the right amount of resources they need. Paradoxically, the

system is gathering a large amount of monitoring data at runtime, which remains unused.

In the ideal abstraction we envision for data scientists, the system is adaptive, able to exploit

monitoring data to learn about workloads, and process user requests into a tailored execution

context. In this work, we study different techniques that have been used to make steps toward

such system awareness, and explore a new way to do so by implementing machine learning

techniques to recommend a specific subset of system configurations for Apache Spark applications.

Furthermore, we present an in depth study of Apache Spark executors configuration, which highlight

the complexity in choosing the best one for a given workload.

Dynamic classifiers for neonatal brain monitoring

Brain injury due to lack of oxygen or impaired blood flow around the time of birth, may cause long term neurological dysfunction or death in severe cases. The treatments need to be initiated as soon as possible and tailored according to the nature of the injury to achieve best outcomes. The Electroencephalogram (EEG) currently provides the best insight into neurological activities. However, its interpretation presents formidable challenge for the neurophsiologists. Moreover, such expertise is not widely available particularly around the clock in a typical busy Neonatal Intensive Care Unit (NICU). Therefore, an automated computerized system for detecting and grading the severity of brain injuries could be of great help for medical staff to diagnose and then initiate on-time treatments. In this study, automated systems for detection of neonatal seizures and grading the severity of Hypoxic-Ischemic Encephalopathy (HIE) using EEG and Heart Rate (HR) signals are presented. It is well known that there is a lot of contextual and temporal information present in the EEG and HR signals if examined at longer time scale. The systems developed in the past, exploited this information either at very early stage of the system without any intelligent block or at very later stage where presence of such information is much reduced. This work has particularly focused on the development of a system that can incorporate the contextual information at the middle (classifier) level. This is achieved by using dynamic classifiers that are able to process the sequences of feature vectors rather than only one feature vector at a time.

Bimodal adaptive hypermedia and interactive multimedia a web-based learning environment based on Kolb's theory of learning style

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.

Unraveling networked learning initiatives: an analytic framework

Networked learning happens naturally within the social systems of which we are all part. However, in certain circumstances individuals may want to actively take initiative to initiate interaction with others they are not yet regularly in exchange with. This may be the case when external influences and societal changes require innovation of existing practices. This paper proposes a framework with relevant dimensions providing insight into precipitated characteristics of designed as well as ‘fostered or grown’ networked learning initiatives. Networked learning initiatives are characterized as “goal-directed, interest-, or needs based activities of a group of (at least three) individuals that initiate interaction across the boundaries of their regular social systems”. The proposed framework is based on two existing research traditions, namely 'networked learning' and 'learning networks', comparing, integrating and building upon knowledge from both perspectives. We uncover some interesting differences between definitions, but also similarities in the way they describe what ‘networked’ means and how learning is conceptualized. We think it is productive to combine both research perspectives, since they both study the process of learning in networks extensively, albeit from different points of view, and their combination can provide valuable insights in networked learning initiatives. We uncover important features of networked learning initiatives, characterize actors and connections of which they are comprised and conditions which facilitate and support them. The resulting framework could be used both for analytic purposes and (partly) as a design framework. In this framework it is acknowledged that not all successful networks have the same characteristics: there is no standard ‘constellation’ of people, roles, rules, tools and artefacts, although there are indications that some network structures work better than others. Interactions of individuals can only be designed and fostered till a certain degree: the type of network and its ‘growth’ (e.g. in terms of the quantity of people involved, or the quality and relevance of co-created concepts, ideas, artefacts and solutions to its ‘inhabitants’) is in the hand of the people involved. Therefore, the framework consists of dimensions on a sliding scale. It introduces a structured and analytic way to look at the precipitation of networked learning initiatives: learning networks. Successive research on the application of this framework and feedback from the networked learning community is needed to further validate it’s usability and value to both research as well as practice.

Recommending peers for learning: Matching on dissimilarity in interpretations to provoke breakdown

People recommenders are a widespread feature of social networking sites and educational social learning platforms alike. However, when these systems are used to extend learners’ Personal Learning Networks, they often fall short of providing recommendations of learning value to their users. This paper proposes a design of a people recommender based on content-based user profiles, and a matching method based on dissimilarity therein. It presents the results of an experiment conducted with curators of the content curation site Scoop.it!, where curators rated personalized recommendations for contacts. The study showed that matching dissimilarity of interpretations of shared interests is more successful in providing positive experiences of breakdown for the curator than is matching on similarity. The main conclusion of this paper is that people recommenders should aim to trigger constructive experiences of breakdown for their users, as the prospect and potential of such experiences encourage learners to connect to their recommended peers.

Unraveling networked learning initiatives: an analytic framework

Networked learning happens naturally within the social systems of which we are all part. However, in certain circumstances individuals may want to actively take initiative to initiate interaction with others they are not yet regularly in exchange with. This may be the case when external influences and societal changes require innovation of existing practices. This paper proposes a framework with relevant dimensions providing insight into precipitated characteristics of designed as well as ‘fostered or grown’ networked learning initiatives. Networked learning initiatives are characterized as “goal-directed, interest-, or needs based activities of a group of (at least three) individuals that initiate interaction across the boundaries of their regular social systems”. The proposed framework is based on two existing research traditions, namely 'networked learning' and 'learning networks', comparing, integrating and building upon knowledge from both perspectives. We uncover some interesting differences between definitions, but also similarities in the way they describe what ‘networked’ means and how learning is conceptualized. We think it is productive to combine both research perspectives, since they both study the process of learning in networks extensively, albeit from different points of view, and their combination can provide valuable insights in networked learning initiatives. We uncover important features of networked learning initiatives, characterize actors and connections of which they are comprised and conditions which facilitate and support them. The resulting framework could be used both for analytic purposes and (partly) as a design framework. In this framework it is acknowledged that not all successful networks have the same characteristics: there is no standard ‘constellation’ of people, roles, rules, tools and artefacts, although there are indications that some network structures work better than others. Interactions of individuals can only be designed and fostered till a certain degree: the type of network and its ‘growth’ (e.g. in terms of the quantity of people involved, or the quality and relevance of co-created concepts, ideas, artefacts and solutions to its ‘inhabitants’) is in the hand of the people involved. Therefore, the framework consists of dimensions on a sliding scale. It introduces a structured and analytic way to look at the precipitation of networked learning initiatives: learning networks. Successive research on the application of this framework and feedback from the networked learning community is needed to further validate it’s usability and value to both research as well as practice.

Comparing architectural styles for distributed expert knowledge modules in intelligent tutoring systems

Intelligent Tutoring Systems (ITSs) are computerized systems for learning-by-doing. These systems provide students with immediate and customized feedback on learning tasks. An ITS typically consists of several modules that are connected to each other. This research focuses on the distribution of the ITS module that provides expert knowledge services. For the distribution of such an expert knowledge module we need to use an architectural style because this gives a standard interface, which increases the reusability and operability of the expert knowledge module. To provide expert knowledge modules in a distributed way we need to answer the research question: ‘How can we compare and evaluate REST, Web services and Plug-in architectural styles for the distribution of the expert knowledge module in an intelligent tutoring system?’. We present an assessment method for selecting an architectural style. Using the assessment method on three architectural styles, we selected the REST architectural style as the style that best supports the distribution of expert knowledge modules. With this assessment method we also analyzed the trade-offs that come with selecting REST. We present a prototype and architectural views based on REST to demonstrate that the assessment method correctly scores REST as an appropriate architectural style for the distribution of expert knowledge modules.

Machine learning classification of surgical pathology reports and chunk recognition for information extraction noise reduction

Background and aims: Machine learning techniques for the text mining of cancer-related clinical documents have not been sufficiently explored. Here some techniques are presented for the pre-processing of free-text breast cancer pathology reports, with the aim of facilitating the extraction of information relevant to cancer staging.

Materials and methods: The first technique was implemented using the freely available software RapidMiner to classify the reports according to their general layout: ‘semi-structured’ and ‘unstructured’. The second technique was developed using the open source language engineering framework GATE and aimed at the prediction of chunks of the report text containing information pertaining to the cancer morphology, the tumour size, its hormone receptor status and the number of positive nodes. The classifiers were trained and tested respectively on sets of 635 and 163 manually classified or annotated reports, from the Northern Ireland Cancer Registry.

Results: The best result of 99.4% accuracy – which included only one semi-structured report predicted as unstructured – was produced by the layout classifier with the k nearest algorithm, using the binary term occurrence word vector type with stopword filter and pruning. For chunk recognition, the best results were found using the PAUM algorithm with the same parameters for all cases, except for the prediction of chunks containing cancer morphology. For semi-structured reports the performance ranged from 0.97 to 0.94 and from 0.92 to 0.83 in precision and recall, while for unstructured reports performance ranged from 0.91 to 0.64 and from 0.68 to 0.41 in precision and recall. Poor results were found when the classifier was trained on semi-structured reports but tested on unstructured.

Conclusions: These results show that it is possible and beneficial to predict the layout of reports and that the accuracy of prediction of which segments of a report may contain certain information is sensitive to the report layout and the type of information sought.

Towards a Resilience Metric Framework for Cyber-Physical Systems

Resilience is widely accepted as a desirable system property for cyber-physical systems. However, there are no metrics that can be used to measure the resilience of cyber-physical systems (CPS) while the multi-dimensional nature of performance in these systems is considered. In this work, we present first results towards a resilience metric framework. The key contributions of this framework are threefold: First, it allows to evaluate resilience with respect to different performance indicators that are of interest. Second, complexities that are relevant to the performance indicators of interest, can be intentionally abstracted. Third and final, it supports the identification of reasons for good or bad resilience to improve system design.

Learning Treewidth-Bounded Bayesian Networks with Thousands of Variables

We present a method for learning treewidth-bounded Bayesian networks from data sets containing thousands of variables. Bounding the treewidth of a Bayesian network greatly reduces the complexity of inferences. Yet, being a global property of the graph, it considerably increases the difficulty of the learning process. Our novel algorithm accomplishes this task, scaling both to large domains and to large treewidths. Our novel approach consistently outperforms the state of the art on experiments with up to thousands of variables.

Person Re-Identification using Deep Convnets with Multi-task Learning

Person re-identification involves recognizing a person across non-overlapping camera views, with different pose, illumination, and camera characteristics. We propose to tackle this problem by training a deep convolutional network to represent a person’s appearance as a low-dimensional feature vector that is invariant to common appearance variations encountered in the re-identification problem. Specifically, a Siamese-network architecture is used to train a feature extraction network using pairs of similar and dissimilar images. We show that use of a novel multi-task learning objective is crucial for regularizing the network parameters in order to prevent over-fitting due to the small size the training dataset. We complement the verification task, which is at the heart of re-identification, by training the network to jointly perform verification, identification, and to recognise attributes related to the clothing and pose of the person in each image. Additionally, we show that our proposed approach performs well even in the challenging cross-dataset scenario, which may better reflect real-world expected performance. 

Emergence and Innovation in Digital Learning: Foundations and Applications

Educational systems worldwide are facing an enormous shift as a result of sociocultural, political, economic, and technological changes. The technologies and practices that have developed over the last decade have been heralded as opportunities to transform both online and traditional education systems. While proponents of these new ideas often postulate that they have the potential to address the educational problems facing both students and institutions and that they could provide an opportunity to rethink the ways that education is organized and enacted, there is little evidence of emerging technologies and practices in use in online education. Because researchers and practitioners interested in these possibilities often reside in various disciplines and academic departments the sharing and dissemination of their work across often rigid boundaries is a formidable task. Contributors to Emergence and Innovation in Digital Learning include individuals who are shaping the future of online learning with their innovative applications and investigations on the impact of issues such as openness, analytics, MOOCs, and social media. Building on work first published in Emerging Technologies in Distance Education, the contributors to this collection harness the dispersed knowledge in online education to provide a one-stop locale for work on emergent approaches in the field. Their conclusions will influence the adoption and success of these approaches to education and will enable researchers and practitioners to conceptualize, critique, and enhance their understanding of the foundations and applications of new technologies.

Learning to recognize faces incrementally

[EN]Most face recognition systems are based on some form of batch learning. Online face recognition is not only more practical, it is also much more biologically plausible. Typical batch learners aim at minimizing both training error and (a measure of) hypothesis complexity. We show that the same minimization can be done incrementally as long as some form of ”scaffolding” is applied throughout the learning process. Scaffolding means: make the system learn from samples that are neither too easy nor too difficult at each step. We note that such learning behavior is also biologically plausible. Experiments using large sequences of facial images support the theoretical claims. The proposed method compares well with other, numerical calculus-based online learners.

Learning to Interpret and Generate Instructional Recipes

Thesis (Ph.D.)--University of Washington, 2016-08