http://repositorio.ufc.br/handle/riufc/384 2026-06-15T13:32:02Z http://repositorio.ufc.br/handle/riufc/86582 Título: Uma abordagem estruturada para observabilidade em microsserviços: taxonomia, catálogo e framework de detecção de anti-padrões Autor(es): Gomes, Francisco Anderson de Almada Abstract: Software systems increasingly rely on microservices-based architectures to enhance scalability, modularity, and continuous deployment. Although this approach simplifies development by promoting a functional decomposition of components, it also introduces significant operational complexity, making failures more frequent and harder to diagnose. In this context, observability emerges as a fundamental concept, defined as the ability to understand and diagnose the internal behavior of a system based on its external outputs, such as metrics, logs, and traces. However, despite its importance, observability is often poorly implemented due to the lack of standardized practices, resulting in ineffective monitoring, alert fatigue, and low efficiency in incident response. Although existing studies discuss concepts, tools, and challenges related to observability, no prior work has focused specifically on observability anti-patterns, recurrent practices that undermine monitoring effectiveness, nor proposed solutions capable of detecting them automatically. Furthermore, there is a lack of a comprehensive taxonomy to classify and organize existing studies on observability. To address these gaps, this thesis presents a taxonomy focused on observability in microservices-based applications, constructed through a systematic mapping of the literature. A total of 84 relevant studies published between 2019 and 2025 were analyzed, providing a comprehensive overview of the field. The review also identifies tools, benchmarking applications, and real-world datasets used in the selected studies. Complementing this contribution, the thesis develops a systematized catalog of observability anti-patterns, offering an approach to identify and mitigate harmful practices. This catalog serves as a practical guide to support teams in building more reliable and efficient systems. In total, 37 anti-patterns were identified, whose relevance was evaluated by 60 experts, achieving an agreement rate of 95%. Finally, the thesis introduces the Observa framework, designed to automatically detect observability anti-patterns. Its operation was evaluated through three experiments and a proof of concept, which demonstrated its technical feasibility and approved its adoption, receiving a recommendation of excellence from the evaluators. Tipo: Tese 2026-01-01T00:00:00Z http://repositorio.ufc.br/handle/riufc/86558 Título: Estudo de problemas de alocação de unidades de patrulhamento em malha urbana: uma abordagem de otimização combinatória Autor(es): Matias, Jhonata Adam Silva Abstract: In this work, we study three optimization problems related to the allocation and routing of patrol units. In the Diameter-Constrained Partitioning problem (DCP), we minimize the number of allocated vehicles while ensuring a maximum response time. We show the equivalence between DCP and the Vertex Coloring Problem (VCP), propose a new formulation for VCP and an exact method to solve this formulation, as well as a post-processing algorithm to improve the balance of patrol areas. Experiments with real instances of DCP showed the efficiency of the exact algorithm in obtaining optimal solutions. In the Min-Max Diameter Partitioning problem (MMDP), we minimize the maximum response time of a given number of vehicles. We show that MMDP is equivalent to the Min-Max Diameter Clustering Problem (MMDCP) and present an exact method for MMDCP. We compare the proposed method with another method from the literature and find that both methods are effective in solving MMDP, with the proposed method offering a more controlled maximum execution time, while the other is faster on average. In the Foot Patrol Problem (FPP), we maximize a weight function on routes that pass through hot segments (street segments with a high crime density). We present a formulation for FPP and use it to develop a column generation-based matheuristic. In experiments, we observed that our heuristic outperformed two other heuristics from the literature in crime coverage. Tipo: Tese 2025-01-01T00:00:00Z http://repositorio.ufc.br/handle/riufc/85962 Título: AETHER: Augmentation and Episodic Task Harnessing for Efficient Recognition Autor(es): Gaspar, Lucas Peres Abstract: Human Activity Recognition (HAR) has become a significant research area for human behavior analysis. Researches from the middle from the last decade prove that deep learning based models are suitable to identify patterns over time series data collected from smart devices (smartphones, smartwatches) and perform accurate activity recognition over a fixed set of observed activities. However, deep learning approaches face some challenges for time series data, like the lack of sufficient data to train an efficient model. Another challenge that comes with HAR is the particularities in the way that users perform the same activity or how the sensor collects the data, generating some individual conditions. This Ph.D. thesis presents a meta-learning algorithm that overcomes the individual condition limitation by providing a training strategy that facilitates the generalization across different tasks, allowing the model to adapt to unseen users, sensors, and activities. It also overcomes the labeled data scarcity limitation by proposing a data augmentation stage to increase the number of observations to be used during the meta-training. The algorithm is compared against the literature using real-world public datasets and obtains encouraging results. The algorithm is compared against the literature using real-world public datasets and obtains good results, surpassing some literature baselines by 20%. Furthermore, the trained meta-models are applied against other public datasets, allowing us to evaluate the meta-models in completely new scenarios, where the proposed algorithm was able to overcome, in some cases, the baselines by over than 40%. Tipo: Tese 2026-01-01T00:00:00Z http://repositorio.ufc.br/handle/riufc/84824 Título: SEPTRON : Predição de sepse usando transformer, ontologia e algoritmo bioinspirado Autor(es): Silva Júnior, Lourival Gerardo da Abstract: The use of Machine Learning (ML) techniques for sepsis prediction has evolved globally, but the selection of clinically relevant variables, essential for building robust predictive models, remains one of the main challenges. This aspect is particularly important for patients with suspected sepsis, who require intensive monitoring during the first hours after admission to Intensive Care Units (ICUs), in order to enable early diagnosis and treatment. In this context, this work proposes the model SEpsis Prediction with TRansformer, Ontology and BioiNspired Algorithm (SEPTRON), built on a Transformer architecture adapted for structured data. The model uses variables initially selected from a conceptual ontological model of sepsis and further refined by a Genetic Algorithm (GA). The contributions include: (i) a structured and replicable pipeline, covering the entire process from data extraction to model evaluation; (ii) creation of the curated dataset MIMIC-IV-EXT-ONTO-SEPSE, derived from the MIMIC-IV database; (iii) Transformer architecture adapted for structured data; and (iv) evidence of the relevance of the chosen clinical variables. The comparative evaluation focused on the performance of the Transformer against four classic models: Logistic Regression, Random Forest, XGBoost, and Long Short-Term Memory (LSTM). Using 4-hour time windows over the first 72 hours of ICU stay, the Transformer with all 17 clinical variables achieved superior performance (AUC-ROC 0.85; recall 0.83). In turn, the SEPTRON, which integrates the Genetic Algorithm to reduce the set of clinical variables from 17 to 8, maintained competitive performance (AUC-ROC 0.83; recall 0.76), combining robustness, interpretability, and computational feasibility. This effectiveness was reinforced by the temporal analysis of the evolution of sepsis probability over the 72 hours. It is concluded, therefore, that SEPTRON represents a relevant proposal to the field of digital health, providing a predictive model capable of anticipating the risk of sepsis in ICU with a good balance between performance, interpretability, and clinical applicability. Tipo: Tese 2025-01-01T00:00:00Z