The object of the study is automated machine learning packages for forecasting time series. The subject of the study is hyperparameter optimization algorithms used in a number of selected packages. The purpose of the work is to compare automated machine learning packages in the context of the problem of forecasting time series and to identify the features of approaches to optimizing hyperparameters used in each package. Research methods: conducting a theoretical analysis of the available literature on the topic of the study, studying the documentation for the automatic machine learning packages involved in the work, conducting experiments, comparing and evaluating the forecasting results using the constructed pipelines, generalizing and interpreting the results. Results of the work: features in the implementation of hyperparameter optimization algorithms for the libraries under consideration are highlighted. / Объект исследования — пакеты автоматизированного машинного обучения для прогнозирования временных рядов. Предмет исследования — алгоритмы оптимизации гиперпараметров применяемые в ряде выбранных пакетов. Цель работы — проведение сравнения пакетов автоматизированного машинного обучения в контексте задачи прогнозирования временных рядов и выявление особенностей подходов к оптимизации гиперпараметров используемых в каждом пакете. Методы исследования: проведение теоретического анализа доступной литературы по теме исследования, изучение документации к задействованным в работе пакетам автоматического машинного обучения, проведение экспериментов, сравнение и оценка результатов прогнозирования с помощью построенных конвейеров, обобщение и интерпретация полученных результатов. Результаты работы: выделены особенности в реализации алгоритмов оптимизации гиперпараметров для рассматриваемых библиотек.

Notwithstanding the recent successes of deep convolutional neural networks for classification tasks, they are sensitive to the selection of their hyperparameters, which impose an exponentially large search space on modern convolutional models. Traditional hyperparameter selection methods include manual, grid, or random search, but these require expert knowledge or are computationally burdensome. Divergently, Bayesian optimization and evolutionary inspired techniques have surfaced as viable alternatives to the hyperparameter problem. Thus, an alternative hybrid approach that combines the advantages of these techniques is proposed. Specifically, the search space is partitioned into discrete-architectural, and continuous and categorical hyperparameter subspaces, which are respectively traversed by a stochastic genetic search, followed by a genetic-Bayesian search. Simulations on a prominent image classification task reveal that the proposed method results in an overall classification accuracy improvement of 0.87% over unoptimized baselines, and a greater than 97% reduction in computational costs compared to a commonly employed brute force approach. / Electrical and Mining Engineering / M. Tech. (Electrical Engineering)

[pt] Este trabalho busca usar o comitê de algoritmos de aprendizado por reforço profundo (deep reinforcement learning) sob uma nova perspectiva. Na literatura, a técnica de comitê é utilizada para melhorar o desempenho, mas, pela primeira vez, esta pesquisa visa utilizar comitê para minimizar a dependência do desempenho de aprendizagem por reforço profundo no ajuste fino de hiperparâmetros, além de tornar o aprendizado mais preciso e robusto. Duas abordagens são pesquisadas; uma considera puramente a agregação de ação, enquanto que a outra também leva em consideração as funções de valor. Na primeira abordagem, é criada uma estrutura de aprendizado online com base no histórico de escolha de ação contínua do comitê com o objetivo de integrar de forma flexível diferentes métodos de ponderação e agregação para as ações dos agentes. Em essência, a estrutura usa o desempenho passado para combinar apenas as ações das melhores políticas. Na segunda abordagem, as políticas são avaliadas usando seu desempenho esperado conforme estimado por suas funções de valor. Especificamente, ponderamos as funções de valor do comitê por sua acurácia esperada, calculada pelo erro da diferença temporal. As funções de valor com menor erro têm maior peso. Para medir a influência do esforço de ajuste do hiperparâmetro, grupos que consistem em uma mistura de diferentes quantidades de algoritmos bem e mal parametrizados foram criados. Para avaliar os métodos, ambientes clássicos como o pêndulo invertido, cart pole e cart pole duplo são usados como benchmarks. Na validação, os ambientes de simulação Half Cheetah v2, um robô bípede, e o Swimmer v2 apresentaram resultados superiores e consistentes demonstrando a capacidade da técnica de comitê em minimizar o esforço necessário para ajustar os hiperparâmetros dos algoritmos. / [en] This work seeks to use ensembles of deep reinforcement learning algorithms from a new perspective. In the literature, the ensemble technique is used to improve performance, but, for the first time, this research aims to use ensembles to minimize the dependence of deep reinforcement learning performance on hyperparameter fine-tuning, in addition to making it more precise and robust. Two approaches are researched; one considers pure action aggregation, while the other also takes the value functions into account. In the first approach, an online learning framework based on the ensemble s continuous action choice history is created, aiming to flexibly integrate different scoring and aggregation methods for the agents actions. In essence, the framework uses past performance to only combine the best policies actions. In the second approach, the policies are evaluated using their expected performance as estimated by their value functions. Specifically, we weigh the ensemble s value functions by their expected accuracy as calculated by the temporal difference error. Value functions with lower error have higher weight. To measure the influence on the hyperparameter tuning effort, groups consisting of a mix of different amounts of well and poorly parameterized algorithms were created. To evaluate the methods, classic environments such as the inverted pendulum, cart pole and double cart pole are used as benchmarks. In validation, the Half Cheetah v2, a biped robot, and Swimmer v2 simulation environments showed superior and consistent results demonstrating the ability of the ensemble technique to minimize the effort needed to tune the the algorithms.

The increasing size and complexity of datasets have accelerated the development of machine learning models and exposed the need for more scalable solutions. This thesis explores challenges associated with large-scale machine learning under data privacy constraints. With the growth of machine learning models, traditional privacy methods such as data anonymization are becoming insufficient. Thus, we delve into alternative approaches, such as differential privacy. Our research addresses the following core areas in the context of scalable privacy-preserving machine learning: First, we examine the implications of data dimensionality on privacy for the application of medical image analysis. We extend the classification algorithm Private Aggregation of Teacher Ensembles (PATE) to deal with high-dimensional labels, and demonstrate that dimensionality reduction can be used to improve privacy. Second, we consider the impact of hyperparameter selection on privacy. Here, we propose a novel adaptive technique for hyperparameter selection in differentially gradient-based optimization. Third, we investigate sampling-based solutions to scale differentially private machine learning to dataset with a large number of records. We study the privacy-enhancing properties of importance sampling, highlighting that it can outperform uniform sub-sampling not only in terms of sample efficiency but also in terms of privacy. The three techniques developed in this thesis improve the scalability of machine learning while ensuring robust privacy protection, and aim to offer solutions for the effective and safe application of machine learning in large datasets. / Den ständigt ökande storleken och komplexiteten hos datamängder har accelererat utvecklingen av maskininlärningsmodeller och gjort behovet av mer skalbara lösningar alltmer uppenbart. Den här avhandlingen utforskar tre utmaningar förknippade med storskalig maskininlärning under dataskyddskrav. För stora och komplexa maskininlärningsmodeller blir traditionella metoder för integritet, såsom datananonymisering, otillräckliga. Vi undersöker därför alternativa tillvägagångssätt, såsom differentiell integritet. Vår forskning behandlar följande utmaningar inom skalbar och integitetsmedveten maskininlärning: För det första undersöker vi hur hög data-dimensionalitet påverkar integriteten för medicinsk bildanalys. Vi utvidgar klassificeringsalgoritmen Private Aggregation of Teacher Ensembles (PATE) för att hantera högdimensionella etiketter och visar att dimensionsreducering kan användas för att förbättra integriteten. För det andra studerar vi hur valet av hyperparametrar påverkar integriteten. Här föreslår vi en ny adaptiv teknik för val av hyperparametrar i gradient-baserad optimering med garantier på differentiell integritet. För det tredje granskar vi urvalsbaserade lösningar för att skala differentiellt privat maskininlärning till stora datamängder. Vi studerar de integritetsförstärkande egenskaperna hos importance sampling och visar att det kan överträffa ett likformigt urval av sampel, inte bara när det gäller effektivitet utan även för integritet. De tre teknikerna som utvecklats i denna avhandling förbättrar skalbarheten för integritetsskyddad maskininlärning och syftar till att erbjuda lösningar för effektiv och säker tillämpning av maskininlärning på stora datamängder. / <p>QC 20231101</p>

An investigation into using attention mechanisms for better feature extraction in wildfire spread prediction models. This research examines the U-net architecture to achieve image segmentation, a process that partitions images by classifying pixels into one of two classes. The deep learning models explored in this research integrate modern deep learning architectures, and techniques used to optimize them. The models are trained on 12 distinct observational variables derived from the Google Earth Engine catalog. Evaluation is conducted with accuracy, Dice coefficient score, ROC-AUC, and F1-score. This research concludes that when augmenting U-net with attention mechanisms, the attention component improves feature suppression and recognition, improving overall performance. Furthermore, employing ensemble modeling reduces bias and variation, leading to more consistent and accurate predictions. When inferencing on wildfire propagation at 30-minute intervals, the architecture presented in this research achieved a ROC-AUC score of 86.2% and an accuracy of 82.1%.

As resource-constrained autonomous vehicles are used for more and more applications, their ability to achieve the lowest possible localization error without expending more power than needed is crucial. Despite this, the parameter settings of the localization systems, both for the platform and the application, are often set arbitrarily. In this thesis, we propose a model-based controller that adapts the parameters of the localization system during run-time by observing conditions in the environment. The test-bed used for experiments consists of maplab, a visual-inertial localization framework, that we execute on the Nvdia Jetson AGX platform. The results show that the linear velocity is the single most important environmental attribute to base the decision of when to update the parameters upon. We also found that while it was not possible to find a direct connection between certain parameters and environmental conditions, a connection could be found between sets of configuration parameters and conditions. Based on these conclusions, we compare model-based controller setups based on three different models: Finite Impulse Response (FIR), AutoRegressive eXogenous input (ARX) and Multi-Layer Perceptron (MLP). The FIR-based controller performed the best. This FIR-based controller is able to select configurations at the appropriate times to keep the error lower than it would be to randomly guess which set of configuration parameters is best. The proposed solution requires offline profiling before it can be implemented on new localization systems, but it can help to reduce the error and power consumption and thus enable more uses of resource-constrained devices. / Användningen av autonoma fordon med begränsade resurser ökar allt mer, vilket i sin tur ökar vikten av att dessa kan lokalisera med lägsta möjliga fel utan att förbruka mer effekt. Trots detta bestäms parametrarna för både hårdvara och i algoritmerna ofta godtyckligt för dessa lokaliseringssystem. I detta examensarbete presenterar vi en lösning till detta, i form av en modellbaserad regulator som anpassar parametrarna baserat på vad den detekterar i omgivningen. Vår testuppställning består av maplab, ett lokaliseringsramverk, som vi exekverar på Nvida Jetson AGX plattformen. Resultaten visar att den linjära hastigheten är den viktigaste miljövariabeln att detektera och använda för att anpassa parametrarna i lokaliseringssystemet. Resultaten visar även att det går att hitta kopplingar mellan konfigurationer och miljövariabler, även om det inte går att hitta mellan specifika konfigurationsparameterar och miljövariabler. Den regulator som presterar bäst visar sig vara en som är baserad på en Finite Impulse Response modell, med en optimeringshorisont på 5 sekunder. Denna presterar bättre än både AutoRegressive eXogenous input baserad regulator och en Multi-Layer Perceptron baserad regulator. Finite Impulse Response regulatorn åstadkommer ett fel som är lägre än slumpmässig gissning, på data den inte sett förut. Lösningen som uppvisas i detta projekt kräver optimering offline för att fungera, men om det utförs kan den reducera både lokaliseringsfelet och effektförbrukningen och genom det skapa nya användningsområden för resursbegränsade enheter.

Исследование продемонстрировало, что методы машинного обучения могут успешно применяться для качественного анализа литературных произведений. Полученные данные свидетельствуют о высокой точности и надежности предложенных подходов. Это открывает широкие возможности для дальнейшего развития и применения данных технологий в филологических исследованиях и литературной критике. / The research demonstrated that machine learning methods can be successfully applied for qualitative analysis of literary works. The obtained data indicate high precision and reliability of the proposed approaches. This paves the way for further development and application of these technologies in philological studies and literary criticism.

Machine learning is spearheading progress for the field of artificial intelligence in terms of providing competition in strategy games to a human opponent, be it in a game of chess, Go or poker. A field of machine learning, which shows the most promising results in playing strategy games, is reinforcement learning. The next milestone for the current research lies in a computer game Starcraft II, which outgrows the previous ones in terms of complexity, and represents a potential new breakthrough in this field. The paper focuses on analysis of the problem, and suggests a solution incorporating a reinforcement learning algorithm A2C and hyperparameter optimization implementation PBT, which could mean a step forward for the current progress.

The diploma thesis is focused on the use of artificial intelligence methods for evaluating the fault condition of machinery. The evaluated data are from a vibrodiagnostic model for simulation of static and dynamic unbalances. The machine learning methods are applied, specifically supervised learning. The thesis describes the Spyder software environment, its alternatives, and the Python programming language, in which the scripts are written. It contains an overview with a description of the libraries (Scikit-learn, SciPy, Pandas ...) and methods — K-Nearest Neighbors (KNN), Support Vector Machines (SVM), Decision Trees (DT) and Random Forests Classifiers (RF). The results of the classification are visualized in the confusion matrix for each method. The appendix includes written scripts for feature engineering, hyperparameter tuning, evaluation of learning success and classification with visualization of the result.

The primary goal of this work is to propose a methodology for discovering hyperparameters. Hyperparameters aid systems in convergence when well-tuned and handcrafted. However, to this end, poorly chosen hyperparameters leave practitioners in limbo, between concerns with implementation or improper choice in hyperparameter and system configuration. We specifically analyze the choice of learning rate in stochastic gradient descent (SGD), a popular algorithm. As a secondary goal, we attempt the discovery of fixed points using smoothing of the loss landscape by exploiting assumptions about its distribution to improve the update rule in SGD. Smoothing of the loss landscape has been shown to make convergence possible in large-scale systems and difficult black-box optimization problems. However, we use stochastic value gradients (SVG) to smooth the loss landscape by learning a surrogate model and then backpropagate through this model to discover fixed points on the real task SGD is trying to solve. Additionally, we construct a gym environment for testing model-free algorithms, such as Proximal Policy Optimization (PPO) as a hyperparameter optimizer for SGD. For tasks, we focus on a toy problem and analyze the convergence of SGD on MNIST using model-free and model-based reinforcement learning methods for control. The model is learned from the parameters of the true optimizer and used specifically for learning rates rather than for prediction. In experiments, we perform in an online and offline setting. In the online setting, we learn a surrogate model alongside the true optimizer, where hyperparameters are tuned in real-time for the true optimizer. In the offline setting, we show that there is more potential in the model-based learning methodology than in the model-free configuration due to this surrogate model that smooths out the loss landscape and makes for more helpful gradients during backpropagation. / L’objectif principal de ce travail est de proposer une méthodologie de découverte des hyperparamètres. Les hyperparamètres aident les systèmes à converger lorsqu’ils sont bien réglés et fabriqués à la main. Cependant, à cette fin, des hyperparamètres mal choisis laissent les praticiens dans l’incertitude, entre soucis de mise en oeuvre ou mauvais choix d’hyperparamètre et de configuration du système. Nous analysons spécifiquement le choix du taux d’apprentissage dans la descente de gradient stochastique (SGD), un algorithme populaire. Comme objectif secondaire, nous tentons de découvrir des points fixes en utilisant le lissage du paysage des pertes en exploitant des hypothèses sur sa distribution pour améliorer la règle de mise à jour dans SGD. Il a été démontré que le lissage du paysage des pertes rend la convergence possible dans les systèmes à grande échelle et les problèmes difficiles d’optimisation de la boîte noire. Cependant, nous utilisons des gradients de valeur stochastiques (SVG) pour lisser le paysage des pertes en apprenant un modèle de substitution, puis rétropropager à travers ce modèle pour découvrir des points fixes sur la tâche réelle que SGD essaie de résoudre. De plus, nous construisons un environnement de gym pour tester des algorithmes sans modèle, tels que Proximal Policy Optimization (PPO) en tant qu’optimiseur d’hyperparamètres pour SGD. Pour les tâches, nous nous concentrons sur un problème de jouet et analysons la convergence de SGD sur MNIST en utilisant des méthodes d’apprentissage par renforcement sans modèle et basées sur un modèle pour le contrôle. Le modèle est appris à partir des paramètres du véritable optimiseur et utilisé spécifiquement pour les taux d’apprentissage plutôt que pour la prédiction. Dans les expériences, nous effectuons dans un cadre en ligne et hors ligne. Dans le cadre en ligne, nous apprenons un modèle de substitution aux côtés du véritable optimiseur, où les hyperparamètres sont réglés en temps réel pour le véritable optimiseur. Dans le cadre hors ligne, nous montrons qu’il y a plus de potentiel dans la méthodologie d’apprentissage basée sur un modèle que dans la configuration sans modèle en raison de ce modèle de substitution qui lisse le paysage des pertes et crée des gradients plus utiles lors de la rétropropagation.