Statistical dialog management for health interventions

Research endeavors on spoken dialogue systems in the 1990s and 2000s have led to the deployment of commercial spoken dialogue systems (SDS) in microdomains such as customer service automation, reservation/booking and question answering systems. Recent research in SDS has been focused on the development of applications in different domains (e.g. virtual counseling, personal coaches, social companions) which requires more sophistication than the previous generation of commercial SDS. The focus of this research project is the delivery of behavior change interventions based on the brief intervention counseling style via spoken dialogue systems. ^ Brief interventions (BI) are evidence-based, short, well structured, one-on-one counseling sessions. Many challenges are involved in delivering BIs to people in need, such as finding the time to administer them in busy doctors' offices, obtaining the extra training that helps staff become comfortable providing these interventions, and managing the cost of delivering the interventions. Fortunately, recent developments in spoken dialogue systems make the development of systems that can deliver brief interventions possible. ^ The overall objective of this research is to develop a data-driven, adaptable dialogue system for brief interventions for problematic drinking behavior, based on reinforcement learning methods. The implications of this research project includes, but are not limited to, assessing the feasibility of delivering structured brief health interventions with a data-driven spoken dialogue system. Furthermore, while the experimental system focuses on harmful alcohol drinking as a target behavior in this project, the produced knowledge and experience may also lead to implementation of similarly structured health interventions and assessments other than the alcohol domain (e.g. obesity, drug use, lack of exercise), using statistical machine learning approaches. ^ In addition to designing a dialog system, the semantic and emotional meanings of user utterances have high impact on interaction. To perform domain specific reasoning and recognize concepts in user utterances, a named-entity recognizer and an ontology are designed and evaluated. To understand affective information conveyed through text, lexicons and sentiment analysis module are developed and tested.^

Learning and Exploiting Camera Geometry for Computer Vision

This work explores the use of statistical methods in describing and estimating camera poses, as well as the information feedback loop between camera pose and object detection. Surging development in robotics and computer vision has pushed the need for algorithms that infer, understand, and utilize information about the position and orientation of the sensor platforms when observing and/or interacting with their environment.

The first contribution of this thesis is the development of a set of statistical tools for representing and estimating the uncertainty in object poses. A distribution for representing the joint uncertainty over multiple object positions and orientations is described, called the mirrored normal-Bingham distribution. This distribution generalizes both the normal distribution in Euclidean space, and the Bingham distribution on the unit hypersphere. It is shown to inherit many of the convenient properties of these special cases: it is the maximum-entropy distribution with fixed second moment, and there is a generalized Laplace approximation whose result is the mirrored normal-Bingham distribution. This distribution and approximation method are demonstrated by deriving the analytical approximation to the wrapped-normal distribution. Further, it is shown how these tools can be used to represent the uncertainty in the result of a bundle adjustment problem.

Another application of these methods is illustrated as part of a novel camera pose estimation algorithm based on object detections. The autocalibration task is formulated as a bundle adjustment problem using prior distributions over the 3D points to enforce the objects' structure and their relationship with the scene geometry. This framework is very flexible and enables the use of off-the-shelf computational tools to solve specialized autocalibration problems. Its performance is evaluated using a pedestrian detector to provide head and foot location observations, and it proves much faster and potentially more accurate than existing methods.

Finally, the information feedback loop between object detection and camera pose estimation is closed by utilizing camera pose information to improve object detection in scenarios with significant perspective warping. Methods are presented that allow the inverse perspective mapping traditionally applied to images to be applied instead to features computed from those images. For the special case of HOG-like features, which are used by many modern object detection systems, these methods are shown to provide substantial performance benefits over unadapted detectors while achieving real-time frame rates, orders of magnitude faster than comparable image warping methods.

The statistical tools and algorithms presented here are especially promising for mobile cameras, providing the ability to autocalibrate and adapt to the camera pose in real time. In addition, these methods have wide-ranging potential applications in diverse areas of computer vision, robotics, and imaging.

Machine Learning-based Techniques to Address Spectral Distortions in Photon Counting X-ray Computed Tomography

Spectral CT using a photon counting x-ray detector (PCXD) shows great potential for measuring material composition based on energy dependent x-ray attenuation. Spectral CT is especially suited for imaging with K-edge contrast agents to address the otherwise limited contrast in soft tissues. We have developed a micro-CT system based on a PCXD. This system enables full spectrum CT in which the energy thresholds of the PCXD are swept to sample the full energy spectrum for each detector element and projection angle. Measurements provided by the PCXD, however, are distorted due to undesirable physical eects in the detector and are very noisy due to photon starvation. In this work, we proposed two methods based on machine learning to address the spectral distortion issue and to improve the material decomposition. This rst approach is to model distortions using an articial neural network (ANN) and compensate for the distortion in a statistical reconstruction. The second approach is to directly correct for the distortion in the projections. Both technique can be done as a calibration process where the neural network can be trained using 3D printed phantoms data to learn the distortion model or the correction model of the spectral distortion. This replaces the need for synchrotron measurements required in conventional technique to derive the distortion model parametrically which could be costly and time consuming. The results demonstrate experimental feasibility and potential advantages of ANN-based distortion modeling and correction for more accurate K-edge imaging with a PCXD. Given the computational eciency with which the ANN can be applied to projection data, the proposed scheme can be readily integrated into existing CT reconstruction pipelines.

Statistical Dialog Management for Health Interventions

Research endeavors on spoken dialogue systems in the 1990s and 2000s have led to the deployment of commercial spoken dialogue systems (SDS) in microdomains such as customer service automation, reservation/booking and question answering systems. Recent research in SDS has been focused on the development of applications in different domains (e.g. virtual counseling, personal coaches, social companions) which requires more sophistication than the previous generation of commercial SDS. The focus of this research project is the delivery of behavior change interventions based on the brief intervention counseling style via spoken dialogue systems. Brief interventions (BI) are evidence-based, short, well structured, one-on-one counseling sessions. Many challenges are involved in delivering BIs to people in need, such as finding the time to administer them in busy doctors' offices, obtaining the extra training that helps staff become comfortable providing these interventions, and managing the cost of delivering the interventions. Fortunately, recent developments in spoken dialogue systems make the development of systems that can deliver brief interventions possible. The overall objective of this research is to develop a data-driven, adaptable dialogue system for brief interventions for problematic drinking behavior, based on reinforcement learning methods. The implications of this research project includes, but are not limited to, assessing the feasibility of delivering structured brief health interventions with a data-driven spoken dialogue system. Furthermore, while the experimental system focuses on harmful alcohol drinking as a target behavior in this project, the produced knowledge and experience may also lead to implementation of similarly structured health interventions and assessments other than the alcohol domain (e.g. obesity, drug use, lack of exercise), using statistical machine learning approaches. In addition to designing a dialog system, the semantic and emotional meanings of user utterances have high impact on interaction. To perform domain specific reasoning and recognize concepts in user utterances, a named-entity recognizer and an ontology are designed and evaluated. To understand affective information conveyed through text, lexicons and sentiment analysis module are developed and tested.

Digital anthropology and educational eGames: learning through behavioural patterns in digital, game-based contexts

The selected publications are focused on the relations between users, eGames and the educational context, and how they interact together, so that both learning and user performance are improved through feedback provision. A key part of this analysis is the identification of behavioural, anthropological patterns, so that users can be clustered based on their actions, and the steps taken in the system (e.g. social network, online community, or virtual campus). In doing so, we can analyse large data sets of information made by a broad user sample,which will provide more accurate statistical reports and readings. Furthermore, this research is focused on how users can be clustered based on individual and group behaviour, so that a personalized support through feedback is provided, and the personal learning process is improved as well as the group interaction. We take inputs from every person and from the group they belong to, cluster the contributions, find behavioural patterns and provide personalized feedback to the individual and the group, based on personal and group findings. And we do all this in the context of educational games integrated in learning communities and learning management systems. To carry out this research we design a set of research questions along the 10-year published work presented in this thesis. We ask if the users can be clustered together based on the inputs provided by them and their groups; if and how these data are useful to improve the learner performance and the group interaction; if and how feedback becomes a useful tool for such pedagogical goal; if and how eGames become a powerful context to deploy the pedagogical methodology and the various research methods and activities that make use of that feedback to encourage learning and interaction; if and how a game design and a learning design must be defined and implemented to achieve these objectives, and to facilitate the productive authoring and integration of eGames in pedagogical contexts and frameworks. We conclude that educational games are a resourceful tool to provide a user experience towards a better personalized learning performance and an enhance group interaction along the way. To do so, eGames, while integrated in an educational context, must follow a specific set of user and technical requirements, so that the playful context supports the pedagogical model underneath. We also conclude that, while playing, users can be clustered based on their personal behaviour and interaction with others, thanks to the pattern identification. Based on this information, a set of recommendations are provided Digital Anthropology and educational eGames 6 /216 to the user and the group in the form of personalized feedback, timely managed for an optimum impact on learning performance and group interaction level. In this research, Digital Anthropology is introduced as a concept at a late stage to provide a backbone across various academic fields including: Social Science, Cognitive Science, Behavioural Science, Educational games and, of course, Technology-enhance learning. Although just recently described as an evolution of traditional anthropology, this approach to digital behaviour and social structure facilitates the understanding amongst fields and a comprehensive view towards a combined approach. This research takes forward the already existing work and published research onusers and eGames for learning, and turns the focus onto the next step — the clustering of users based on their behaviour and offering proper, personalized feedback to the user based on that clustering, rather than just on isolated inputs from every user. Indeed, this pattern recognition in the described context of eGames in educational contexts, and towards the presented aim of personalized counselling to the user and the group through feedback, is something that has not been accomplished before.

Performance Evaluation of a Statistical and a Neural Network Model for Nonrigid Shape-Based Registration

Shape-based registration methods frequently encounters in the domains of computer vision, image processing and medical imaging. The registration problem is to find an optimal transformation/mapping between sets of rigid or nonrigid objects and to automatically solve for correspondences. In this paper we present a comparison of two different probabilistic methods, the entropy and the growing neural gas network (GNG), as general feature-based registration algorithms. Using entropy shape modelling is performed by connecting the point sets with the highest probability of curvature information, while with GNG the points sets are connected using nearest-neighbour relationships derived from competitive hebbian learning. In order to compare performances we use different levels of shape deformation starting with a simple shape 2D MRI brain ventricles and moving to more complicated shapes like hands. Results both quantitatively and qualitatively are given for both sets.

Virtual Reality as a tool to Assist Learning in Aerospace Assembly

This study examines whether virtual reality (VR) is more superior to paper-based instructions in increasing the speed at which individuals learn a new assembly task. Specifically, the work seeks to quantify any learning benefits when individuals have been given the opportunity and compares the performance of two groups using virtual and hardcopy media types to pre-learn the task. A build experiment based on multiple builds of an aircraft panel showed that a group of people who pre-learned the assembly task using a VR environment completed their builds faster (average build time 29.5% lower). The VR group also made fewer references to instructional materials (average number of references 38% lower) and made fewer errors than a group using more traditional, hard copy instructions. These outcomes were more pronounced during build one with differences in build time and number of references showing limited statistical differences.

Machine learning and data decompositions for complex networked dynamical systems

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

Essays on Machine Learning and Hedonic Models

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

Predominant learning styles among pharmacy students at the Federal University of Paraná, Brazil

Background: Learning styles are cognitive, emotional, and physiological traits, as well as indicators of how learners perceive, interact, and respond to their learning environments. According to Honey-Mumford, learning styles are classified as active, reflexive, theoretical, and pragmatic. Objective: The purpose of this study was to identify the predominant learning styles among pharmacy students at the Federal University of Paraná, Brazil. Methods: An observational, cross-sectional, and descriptive study was conducted using the Honey-Alonso Learning Style Questionnaire. Students in the Bachelor of Pharmacy program were invited to participate in this study. The questionnaire comprised 80 randomized questions, 20 for each of the four learning styles. The maximum possible score was 20 points for each learning style, and cumulative scores indicated the predominant learning styles among the participants. Honey-Mumford (1986) proposed five preference levels for each style (very low, low, moderate, high, and very high), called a general interpretation scale, to avoid student identification with one learning style and ignoring the characteristics of the other styles. Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) version 20.0. Results: This study included 297 students (70% of all pharmacy students at the time) with a median age of 21 years old. Women comprised 77.1% of participants. The predominant style among pharmacy students at the Federal University of Paraná was the pragmatist, with a median of 14 (high preference). The pragmatist style prevails in people who are able to discover techniques related to their daily learning because such people are curious to discover new strategies and attempt to verify whether the strategies are efficient and valid. Because these people are direct and objective in their actions, pragmatists prefer to focus on practical issues that are validated and on problem situations. There was no statistically significant difference between genders with regard to learning styles. Conclusion: The pragmatist style is the prevailing style among pharmacy students at the Federal University of Paraná. Although students may have a learning preference that preference is not the only manner in which students can learn, neither their preference is the only manner in which students can be taught. Awareness of students’ learning styles can be used to adapt the methodology used by teachers to render the teaching-learning process effective and long lasting. The content taught to students should be presented in different manners because varying teaching methods can develop learning skills in students.

The Effectiveness of Outdoor Education on Environmental Learning, Appreciation, and Activism

The main objective of this research was to determine the effectiveness of outdoor education on student knowledge retention, appreciation for nature, and environmental activism in a college level course on south Florida ecology. Six class sections were given quizzes on four course topics either post-lecture or post-field trip. Students were also given pre-course and post-course opinion surveys. Although mean quiz scores for the post-field trip were higher than for the post-lecture, statistical analysis determined that there was no significant difference in quiz scores for location taken (post-lecture or post-field trip). Survey results show a correlation between knowledge of environmental issues and environmental activism. Even though student survey responses point to outdoor education and field trips being the most effective method of learning and influential on appreciation for nature, the quiz scores do not reflect such.

Deep Learning Architectures for time of arrival detection in Acoustic Emissions Monitoring

Acoustic Emission (AE) monitoring can be used to detect the presence of damage as well as determine its location in Structural Health Monitoring (SHM) applications. Information on the time difference of the signal generated by the damage event arriving at different sensors is essential in performing localization. This makes the time of arrival (ToA) an important piece of information to retrieve from the AE signal. Generally, this is determined using statistical methods such as the Akaike Information Criterion (AIC) which is particularly prone to errors in the presence of noise. And given that the structures of interest are surrounded with harsh environments, a way to accurately estimate the arrival time in such noisy scenarios is of particular interest. In this work, two new methods are presented to estimate the arrival times of AE signals which are based on Machine Learning. Inspired by great results in the field, two models are presented which are Deep Learning models - a subset of machine learning. They are based on Convolutional Neural Network (CNN) and Capsule Neural Network (CapsNet). The primary advantage of such models is that they do not require the user to pre-define selected features but only require raw data to be given and the models establish non-linear relationships between the inputs and outputs. The performance of the models is evaluated using AE signals generated by a custom ray-tracing algorithm by propagating them on an aluminium plate and compared to AIC. It was found that the relative error in estimation on the test set was < 5% for the models compared to around 45% of AIC. The testing process was further continued by preparing an experimental setup and acquiring real AE signals to test on. Similar performances were observed where the two models not only outperform AIC by more than a magnitude in their average errors but also they were shown to be a lot more robust as compared to AIC which fails in the presence of noise.

Deep learning models for the automatic analysis of volleyball plays

Collecting and analysing data is an important element in any field of human activity and research. Even in sports, collecting and analyzing statistical data is attracting a growing interest. Some exemplar use cases are: improvement of technical/tactical aspects for team coaches, definition of game strategies based on the opposite team play or evaluation of the performance of players. Other advantages are related to taking more precise and impartial judgment in referee decisions: a wrong decision can change the outcomes of important matches. Finally, it can be useful to provide better representations and graphic effects that make the game more engaging for the audience during the match. Nowadays it is possible to delegate this type of task to automatic software systems that can use cameras or even hardware sensors to collect images or data and process them. One of the most efficient methods to collect data is to process the video images of the sporting event through mixed techniques concerning machine learning applied to computer vision. As in other domains in which computer vision can be applied, the main tasks in sports are related to object detection, player tracking, and to the pose estimation of athletes. The goal of the present thesis is to apply different models of CNNs to analyze volleyball matches. Starting from video frames of a volleyball match, we reproduce a bird's eye view of the playing court where all the players are projected, reporting also for each player the type of action she/he is performing.

Interdisciplinary project-based learning: technology for improving student cognition

The article studies a way of enhancing student cognition by using interdisciplinary project-based learning (IPBL) in a higher education institution. IPBL is a creative pedagogic approach allowing students of one area of specialisation to develop projects for students with different academic profiles. The application of this approach in the Ural State University of Economics resulted in a computer-assisted learning system (CALS) designed by IT students. The CALS was used in an analytical chemistry course with students majoring in Commodities Management and Expertise (‘expert’ students). To test how effective the technology was, the control and experimental groups were formed. In the control group, learning was done with traditional methods. In the experimental group, it was reinforced by IPBL. A statistical analysis of the results, with an application of Pearson χ 2 test, showed that the cognitive levels in both IT and ‘expert’ experimental groups improved as compared with the control groups. The findings demonstrated that IPBL can significantly enhance learning. It can be implemented in any institution of higher or secondary education that promotes learning, including the CALS development and its use for solving problems in different subject areas.

Concepts and methods for efficient decentralized learning in federated settings

Deep Neural Networks (DNNs) have revolutionized a wide range of applications beyond traditional machine learning and artificial intelligence fields, e.g., computer vision, healthcare, natural language processing and others. At the same time, edge devices have become central in our society, generating an unprecedented amount of data which could be used to train data-hungry models such as DNNs. However, the potentially sensitive or confidential nature of gathered data poses privacy concerns when storing and processing them in centralized locations. To this purpose, decentralized learning decouples model training from the need of directly accessing raw data, by alternating on-device training and periodic communications. The ability of distilling knowledge from decentralized data, however, comes at the cost of facing more challenging learning settings, such as coping with heterogeneous hardware and network connectivity, statistical diversity of data, and ensuring verifiable privacy guarantees. This Thesis proposes an extensive overview of decentralized learning literature, including a novel taxonomy and a detailed description of the most relevant system-level contributions in the related literature for privacy, communication efficiency, data and system heterogeneity, and poisoning defense. Next, this Thesis presents the design of an original solution to tackle communication efficiency and system heterogeneity, and empirically evaluates it on federated settings. For communication efficiency, an original method, specifically designed for Convolutional Neural Networks, is also described and evaluated against the state-of-the-art. Furthermore, this Thesis provides an in-depth review of recently proposed methods to tackle the performance degradation introduced by data heterogeneity, followed by empirical evaluations on challenging data distributions, highlighting strengths and possible weaknesses of the considered solutions. Finally, this Thesis presents a novel perspective on the usage of Knowledge Distillation as a mean for optimizing decentralized learning systems in settings characterized by data heterogeneity or system heterogeneity. Our vision on relevant future research directions close the manuscript.