[pt] Situada no campo da filosofia da educação, a tese dialoga também com o campo das tecnologias educacionais. O trabalho busca uma compreensão filosófica dos impactos da aprendizagem de máquina ou machine learning na educação. Para isso, dedica-se aos pressupostos subjacentes à aprendizagem de máquina em articulação com os pressupostos subjacentes à concepção de aprendizagem humana que descende do enativismo. Defende-se que a chegada da aprendizagem de máquina na educação encontra um campo em que ainda predomina o paradigma cognitivista, o qual é bastante profícuo para que germinem as tecnologias baseadas em dados e redes neurais. Avança-se para demonstrar que esse paradigma, no entanto, vem sendo desafiado por outras abordagens de pesquisa que se dedicam à mente humana, dentre as quais se destaca o enativismo. São explicitadas as bases teóricas fundamentais do enativismo, e como elas se desdobram em pressupostos para uma aprendizagem humana que é corporificada e essencialmente orientada ao acoplamento do ser com o mundo e com os outros agentes. É dedicada atenção especial aos impactos da aprendizagem de máquina na autonomia do cognoscente, a qual, sob a perspectiva do enativismo, somente pode existir e se manter nas trocas com o meio e com aqueles que habitam e formam esse ambiente. Demonstra-se que, para que as tecnologias algorítmicas sejam adequadas a uma concepção de cognição e de aprendizagem enativista, é preciso buscar um caminho de valorização ainda maior do corpo na aprendizagem, bem como da intersubjetividade, uma vez que as relações entre os agentes cognitivos não são concebidas como articulações opcionais, mas como um elemento que está no cerne da atividade cognitiva humana e do qual essa atividade emerge. / [en] This work is situated in the field of philosophy of education, and also relates to the field of educational technologies. The thesis seeks a philosophical understanding of the impacts of machine learning in education. To do so, it addresses the assumptions underlying machine learning in conjunction with the premises underlying the conception of human learning that derive from enactivism. It is argued that the arrival of machine learning in education found fertile ground in which the cognitivist paradigm still prevails, a situation that is rather fruitful for technologies based on data and neural networks to thrive. The thesis demonstrates that this paradigm, however, has been challenged by other research approaches that are dedicated to the human mind, among which enactivism is emphasized. The fundamental theoretical underpinnings of autopoietic enactivism are explained, as well as how they unfold in assumptions for a notion of human learning that is embodied and essentially oriented towards the coupling of the human organism with the world and with other agents. Particular attention is drawn to the impacts of machine learning on the autonomy of the cognizer, which, from the perspective of enactivism, can only exist and be maintained in exchanges with the environment and with those who inhabit and shape this environment. It is shown that for algorithmic technologies to be suited to an enactivist conception of cognition and learning a greater appreciation of the body in learning is necessary, as well as intersubjectivity, since the connections between cognitive agents are not conceived as optional articulations, but as an element that is at the core of human cognitive activity and from which this activity emerges.

Parkinson's Disease is characterized by the loss of dopaminergic neurons in the substantia nigra pars compacta (SNc). The SNc supplies the basal ganglia (BG) via dopaminergic projections which innervate D1 and D2 receptors that mediate motor control. The BG also mediates cognitive processes and eye movement, parallel to its involvement in motor control. Behavioural correlates of PD have been established from previous countermanding tasks and population neural activity has been shown to correlate with PD disease state, but a reliable means to find patient-specific biomarkers of disease remains unknown. Here, we propose using eye movements and electroencephalography (EEG) to capture neural correlates of dysfunction in PD. We have developed a novel saccade-based stop-signal task in VR that probes the subject's ability to recruit the neural processes involved in action selection and response inhibition. We have tested this system on 7 healthy subjects and verified that we could identify key signature changes in the EEG profile during left and right saccade, countermand, and antisaccades similar to those found in similar reach tasks. The successful completion of a countermand (revoking a planned action) stop trial requires large synchronization of frontal theta and motor beta activity, representing the BG-thalamocortical loop recruiting the necessary processes to inhibit motor responses. The pattern in the event-related potentials that illustrates this is a strong event-related synchronization (ERS) peak followed by an event-related desynchronization (ERD) dip, and increased weights in the scalp topology at the frontal-parietal region. Since tasks involving response inhibition serve to probe the subject’s ability to revoke a planned action, it does not matter whether the task was completed using hand movements or saccades. Our ERP isolated from Independent Component Analysis (ICA) resembles the ERP from previous literature, and exhibits increased weights on the sensorimotor region with a narrow band beta. This narrow band beta range is subject-specific and can be better visualized by using a modelling approach called FOOOF (fitting oscillations one over f). Lastly, the increased decoding performance in each subject's successive recording session suggests that using subject-specific features positively biases the model towards enhanced generalizability. Our experimental platform provides a robust framework that accounts for trial-by-trial variability, and can capture the presence of and evoke beta oscillations in healthy subjects.

The art of anomaly detection is a relevant topic for most producing companies since it allows for real-time quality assurance in production. However, previous research is lacking on the applicability of anomaly detection methods on non-synthetic image datasets. Using a dataset provided by Swedish Match consisting of 943 images of snus cans without lids, we offer an extension to a recent anomaly detection benchmark study by assessing how 29 anomaly detection algorithms perform on our non-synthetic dataset. The results showed that fully supervised methods performed the best, and that labelled data significantly improved model performance. Although the achieved results were not satisfactory in terms of AUCROC and AUCPR, there were clear indications that performance can be improved by increasing the amount of training data. The best-performing model was Logistic Regression. / Avvikelsedetektering är ett relevant ämne för de flesta aktörerna inom tillverkningsindustrin eftersom det möjliggör kvalitetssäkring i realtid i produktionskedjor. I tidigare forskning har det saknats studier gjorda med verklighetstrogna, icke-syntetiska dataset. Med hjälp av ett dataset tillhandahållet av Swedish Match bestående av 943 bilder på öppna snusdosor tillför vi en vetenskaplig påbyggnad till en nyligen publicerad jämförelsestudie inom avvikelsedetektering. Detta genom att träna och utvärdera 29 avvikelsedetekteringsmodeller på vårt icke-syntetiska dataset. Resultaten visade att fully supervised-modellerna presterade bäst, och att klassificerad träningsdata ökar prestandan. Trots att modellerna generellt uppnådde låg AUCPR och AUCROC finns det tydliga indikationer på att detta är uppnåbart genom att utöka träningsdatamängden. Den bäst presterande modellen var Logistic Regression.

To solve real-world problems, machines are required to perceive multiple modalities and fuse the information from them. This thesis studies learning to understand and fuse multimodal information. Existing approaches follow a three-stage learning paradigm. The first stage is to train models for each modality. This process for video understanding models is usually based on supervised training, which is not scalable. Moreover, these modality-specific models are updated rather frequently nowadays with improving single-modality perception abilities. The second stage is crossmodal pretraining, which trains a model to align and fuse multiple modalities based on paired multimodal data, such as video-caption pairs. This process is resource-consuming and expensive. The third stage is to further fine-tune or prompt the resulting model from the second stage towards certain downstream tasks. The key bottleneck of conventional methods lies in the continuous feature representation used for non-textual modalities, which is usually costly to align and fuse with text. In this thesis, we investigate the representation and the fusion based on textual concepts. We propose to map non-textual modalities to textual concepts and then fuse these textual concepts using text models. We systematically study various specific methods of mapping and different architectures for fusion. The proposed methods include an end-to-end video-based text generation model with differentiable tokenization for video and audio concepts, a contrastive-model-based architecture with zero-shot concept extractor, a deep concept injection algorithm enabling language models to solve multimodal tasks without any training, and a distant supervision framework learning concepts in a long temporal span. With our concept representation, we empirically demonstrate that without several orders of magnitude more cost for the crossmodal pretraining stage, our models are able to achieve competitive or even superior performance on downstream tasks such as video question answering, video captioning, text-video retrieval, and audio-video dialogue. We also examine the possible limitations of concept representations such as when the text quality of a dataset is poor. We believe we show a potential path towards upgradable multimodal intelligence, whose components can be easily updated towards new models or new modalities of data.

This research methodology isolates coding properties and identifies the probability of security vulnerabilities using machine learning and historical data. Several approaches characterize the effectiveness of detecting security-related bugs that manifest as vulnerabilities, but none utilize vulnerability patch information. The main contribution of this research is a framework to analyze LLVM Intermediate Representation Code and merging core source code representations using source code properties. This research is beneficial because it allows source programs to be transformed into a graphical form and users can extract specific code properties related to vulnerable functions. The result is an improved approach to detect, identify, and track software system vulnerabilities based on a performance evaluation. The methodology uses historical function level vulnerability information, unique feature extraction techniques, a novel code property graph, and learning algorithms to minimize the amount of end user domain knowledge necessary to detect vulnerabilities in applications. The analysis shows approximately 99% precision and recall to detect known vulnerabilities in the National Institute of Standards and Technology (NIST) Software Assurance Metrics and Tool Evaluation (SAMATE) project. Furthermore, 72% percent of the historical vulnerabilities in the OpenSSL testing environment were detected using a linear support vector classifier (SVC) model.

Although its applications have spread beyond computer science field, the process of machine learning still has some challenges for both expert and novice users. Machine learning platform aims to automate and accelerate the delivery cycle of using machine learning techniques. The objective of this degree project is to generate a user-centred design for an input interface of a machine-learning platform. To answer the research question, there are three methods conducted sequentially: 1) interviews; 2) prototyping; and 3) design evaluation. From the initial interview, we concluded users’ problems and expectations into 11 initial design requirements that should be incorporated into our future platform. The prototype testing focused on checking and improving the functionalities, rather than the visual appearance of the product. Finally, in the design evaluation method, the research delivered design recommendations consisting of five implications: 1) start with a clear definition of the specific machine learning goal; 2) present states of machine learning with a straight-forward flow that promotes learning-opportunity; 3) enable two-way transitions between all states; 4) accommodate different users’ goals with multiple scenarios; and 5) provide expert users with more control to customize the models. / Trots att dess tillämpningar har spridit sig utöver datavetenskapliga fält, behöver utvecklingen av framgångsrik användning av maskininlärning fortfarande anspråkiga komplexa metoder. Maskininlärningsplattform syftar till att automatisera och påskynda leveranscykeln för att använda maskininlärningstekniker. Syftet med detta examensarbete är att generera en användarcentrerad design för ett ingångsgränssnitt för en maskininlärningsplattform. För att besvara forskningsfrågan finns det tre metoder som genomförs i följd: 1) intervjuer; 2) prototypning; och 3) designutvärdering. Från den första intervjun avslutade vi användarnas problem och förväntningar i 11 ursprungliga designkrav som bör integreras av vår framtida plattform. Prototyptesten fokuserade på att kontrollera och förbättra funktionaliteterna snarare än det visuella utseendet på produkten. Avslutningsvis, i designbedömningsmetoden, levererade forskningen designrekommendationer bestående av fem implikationer: 1) börja med en tydlig definition av maskininlärningsmålet; 2) nuvarande stater med ett rakt framåtflöde som främjar inlärningsmöjligheter; 3) möjliggöra tvåvägsövergångar mellan tillstånd; 4) Rymma olika användares mål med flera scenarier; och 5) ge experter användare mer kontroll.

A problem that exhibition halls have is the balance between having good indoor air quality andminimizing energy waste due to the naturally slow decrease of CO2 concentration, which causes Heat-ing, Ventilation and Air-Conditioning systems to keep ventilating empty halls when occupants have leftthe vicinity. Several studies have been made on the topic of CO2 prediction and occupancy predictionbased on CO2 for smaller spaces such as offices and schools. However, few studies have been madefor bigger venues where a larger group of people gather. An online machine learning model using theRiver library was developed to tackle this problem by predicting the CO2 ahead of time. Five datasetswere used for training and predicting, three with real data and two with simulated data. The resultsfrom this model was compared with three already developed traditional models in order to evaluate theperformance of an online machine learning model compared to traditional models. The online machinelearning model was successful in predicting CO2 one hour ahead of time considerably faster than thetraditional models, achieving a r2 score of up to 0.95.

Chemical engineers have routinely used computational tools for modeling, optimizing, and debottlenecking chemical processes. Because of the advances in computational science over the past decade, multivariate statistics and machine learning have become an integral part of the computerization of chemical processes. In this research, we look into using multivariate statistics, machine learning tools, and their combinations through a series of case studies including a case with a successful industrial deployment of machine learning models for fermentation. We use both commercially-available software tools, Aspen ProMV and Python, to demonstrate the feasibility of the computational tools. This work demonstrates a novel application of ensemble-based machine learning methods in bioprocessing, particularly for the prediction of different fermenter types in a fermentation process (to allow for successful data integration) and the prediction of the onset of foaming. We apply two ensemble frameworks, Extreme Gradient Boosting (XGBoost) and Random Forest (RF), to build classification and regression models. Excessive foaming can interfere with the mixing of reactants and lead to problems, such as decreasing effective reactor volume, microbial contamination, product loss, and increased reaction time. Physical modeling of foaming is an arduous process as it requires estimation of foam height, which is dynamic in nature and varies for different processes. In addition to foaming prediction, we extend our work to control and prevent foaming by allowing data-driven ad hoc addition of antifoam using exhaust differential pressure as an indicator of foaming. We use large-scale real fermentation data for six different types of sporulating microorganisms to predict foaming over multiple strains of microorganisms and build exploratory time-series driven antifoam profiles for four different fermenter types. In order to successfully predict the antifoam addition from the large-scale multivariate dataset (about half a million instances for 163 batches), we use TPOT (Tree-based Pipeline Optimization Tool), an automated genetic programming algorithm, to find the best pipeline from 600 other pipelines. Our antifoam profiles are able to decrease hourly volume retention by over 53% for a specific fermenter. A decrease in hourly volume retention leads to an increase in fermentation product yield. We also study two different cases associated with the manufacturing of polyolefins, particularly LDPE (low-density polyethylene) and HDPE (high-density polyethylene). Through these cases, we showcase the usage of machine learning and multivariate statistical tools to improve process understanding and enhance the predictive capability for process optimization. By using indirect measurements such as temperature profiles, we demonstrate the viability of such measures in the prediction of polyolefin quality parameters, anomaly detection, and statistical monitoring and control of the chemical processes associated with a LDPE plant. We use dimensionality reduction, visualization tools, and regression analysis to achieve our goals. Using advanced analytical tools and a combination of algorithms such as PCA (Principal Component Analysis), PLS (Partial Least Squares), Random Forest, etc., we identify predictive models that can be used to create inferential schemes. Soft-sensors are widely used for on-line monitoring and real-time prediction of process variables. In one of our cases, we use advanced machine learning algorithms to predict the polymer melt index, which is crucial in determining the product quality of polymers. We use real industrial data from one of the leading chemical engineering companies in the Asia-Pacific region to build a predictive model for a HDPE plant. Lastly, we show an end-to-end workflow for deep learning on both industrial and simulated polyolefin datasets. Thus, using these five cases, we explore the usage of advanced machine learning and multivariate statistical techniques in the optimization of chemical and biochemical processes. The recent advances in computational hardware allow engineers to design such data-driven models, which enhances their capacity to effectively and efficiently monitor and control a process. We showcase that even non-expert chemical engineers can implement such machine learning algorithms with ease using open-source or commercially available software tools. / Doctor of Philosophy / Most chemical and biochemical processes are equipped with advanced probes and connectivity sensors that collect large amounts of data on a daily basis. It is critical to manage and utilize the significant amount of data collected from the start and throughout the development and manufacturing cycle. Chemical engineers have routinely used computational tools for modeling, designing, optimizing, debottlenecking, and troubleshooting chemical processes. Herein, we present different applications of machine learning and multivariate statistics using industrial datasets. This dissertation also includes a deployed industrial solution to mitigate foaming in commercial fermentation reactors as a proof-of-concept (PoC). Our antifoam profiles are able to decrease volume loss by over 53% for a specific fermenter. Throughout this dissertation, we demonstrate applications of several techniques like ensemble methods, automated machine learning, exploratory time series, and deep learning for solving industrial problems. Our aim is to bridge the gap from industrial data acquisition to finding meaningful insights for process optimization.

While the toolset known as Machine Learning (ML) is not new, several of the tools available within the toolset have seen revitalization with improved hardware, and have been applied across several domains in the last two decades. Deep Neural Network (DNN) applications have contributed to significant research within Radio Frequency (RF) problems over the last decade, spurred by results in image and audio processing. Machine Learning (ML), and Deep Learning (DL) specifically, are driven by access to relevant data during the training phase of the application due to the learned feature sets that are derived from vast amounts of similar data. Despite this critical reliance on data, the literature provides insufficient answers on how to quantify the data training needs of an application in order to achieve a desired performance. This dissertation first aims to create a practical definition that bounds the problem space of Radio Frequency Machine Learning (RFML), which we take to mean the application of Machine Learning (ML) as close to the sampled baseband signal directly after digitization as is possible, while allowing for preprocessing when reasonably defined and justified. After constraining the problem to the Radio Frequency Machine Learning (RFML) domain space, an understanding of what kinds of Machine Learning (ML) have been applied as well as the techniques that have shown benefits will be reviewed from the literature. With the problem space defined and the trends in the literature examined, the next goal aims at providing a better understanding for the concept of data quality through quantification. This quantification helps explain how the quality of data: affects Machine Learning (ML) systems with regard to final performance, drives required data observation quantity within that space, and impacts can be generalized and contrasted. With the understanding of how data quality and quantity can affect the performance of a system in the Radio Frequency Machine Learning (RFML) space, an examination of the data generation techniques and realizations from conceptual through real-time hardware implementations are discussed. Consequently, the results of this dissertation provide a foundation for estimating the investment required to realize a performance goal within a Deep Learning (DL) framework as well as a rough order of magnitude for common goals within the Radio Frequency Machine Learning (RFML) problem space. / Doctor of Philosophy / Machine Learning (ML) is a powerful toolset capable of solving difficult problems across many domains. A fundamental part of this toolset is the representative data used to train a system. Unlike the domains of image or audio processing, for which datasets are constantly being developed thanks to usage agreements with entities such as Facebook, Google, and Amazon, the field of Machine Learning (ML) within the Radio Frequency (RF) domain, or Radio Frequency Machine Learning (RFML), does not have access to such crowdsourcing means of creating labeled datasets. Therefore data within the Radio Frequency Machine Learning (RFML) problem space must be intentionally cultivated to address the target problem. This dissertation explains the problem space of Radio Frequency Machine Learning (RFML) and then quantifies the effect of quality on data used during the training of Radio Frequency Machine Learning (RFML) systems. Taking this one step further, the work then goes on to provide a means of estimating data quantity needs to achieve high levels of performance based on the current Deep Learning (DL) approach to solve the problem, which in turn can be used as guidance to better refine the approach when the real-world data quantity requirements exceed practical acquisition levels. Finally, the problem of data generation is examined and provides context for the difficulties associated with procuring high quality data for problems in the Radio Frequency Machine Learning (RFML) space.

As research in Artificial Intelligence (AI) advances, it is crucial to focus on having seamless communication between humans and machines in order to effectively accomplish tasks. Smooth human-machine communication requires the machine to be sensible and human-like while interacting with humans, while simultaneously being capable of extracting the maximum information it needs to accomplish the desired task. Since a lot of the tasks required to be solved by machines today involve the understanding of images, training machines to have human-like and effective image-grounded conversations with humans is one important step towards achieving this goal. Although we now have agents that can answer questions asked for images, they are prone to failure from confusing input, and cannot ask clarification questions, in turn, to extract the desired information from humans. Hence, as a first step, we direct our efforts towards making Visual Question Answering agents human-like by making them resilient to confusing inputs that otherwise do not confuse humans. Not only is it crucial for a machine to answer questions reasonably, it should also know how to ask questions sequentially to extract the desired information it needs from a human. Hence, we introduce a novel game called the Visual 20 Questions Game, where a machine tries to figure out a secret image a human has picked by having a natural language conversation with the human. Using deep learning techniques like recurrent neural networks and sequence-to-sequence learning, we demonstrate scalable and reasonable performances on both the tasks. / Master of Science