Abstract
Hypertension is the primary risk factor for cerebral small vessel disease (CSVD). However, its mechanistic links are yet to be completely understood. Advancements in diffusion-weighted magnetic resonance imaging (dMRI) increased sensitivity in detecting subtle white matter (WM) structural integrity changes. 44 hypertension patients without symptomatic CSVD underwent multi-modal evaluation of cerebral structure and function, including dMRI, neuropsychological tests and transcranial Doppler monitoring of the right middle cerebral artery (MCA) and left posterior cerebral artery (PCA) to assess neurovascular coupling (NVC). In the PCA, the modeled NVC curve was studied. We examined the cross-sectional relationship of WM integrity with NVC and cognitive performance, using correlational tractography. Diffusion measures from two dMRI models were used: fractional anisotropy, mean diffusivity, axial diffusivity, and radial diffusivity from diffusion tensor imaging, and quantitative anisotropy (QA) and isotropy from q-space diffeomorphic reconstruction. Regarding the NVC in the PCA, vascular elastic properties and initial response speed markers indicated better functional hyperemia with better WM integrity. However, the amplitude suggested increased NVC with worse WM integrity. In the MCA, increased NVC was associated with lower WM integrity. Better cognitive performance associated with preserved WM integrity. Increased functional hyperemia despite worse WM integrity may reflect less efficient NVC in hypertensive patients, potentially arising from (mal)adaptive mechanisms and brain network reorganization in response to CSVD. This observational study highlights the potential of transcranial Doppler and QA as susceptibility markers of pre-symptomatic CSVD.

Abstract
Semantic segmentation consists of predicting a semantic label for each image pixel. While existing deep learning approaches achieve high accuracy, they often overlook the ordinal relationships between classes, which can provide critical domain knowledge (e.g., the pupil lies within the iris, and lane markings are part of the road). This paper introduces novel methods for spatial ordinal segmentation that explicitly incorporate these inter-class dependencies. By treating each pixel as part of a structured image space rather than as an independent observation, we propose two regularization terms and a new metric to enforce ordinal consistency between neighboring pixels. Two loss regularization terms and one metric are proposed for structural ordinal segmentation, which penalizes predictions of non-ordinal adjacent classes. Five biomedical datasets and multiple configurations of autonomous driving datasets demonstrate the efficacy of the proposed methods. Our approach achieves improvements in ordinal metrics and enhances generalization, with up to a 15.7% relative increase in the Dice coefficient. Importantly, these benefits come without additional inference time costs. This work highlights the significance of spatial ordinal relationships in semantic segmentation and provides a foundation for further exploration in structured image representations.

Abstract
Mutation testing maybe used to guide test case generation and as a technique to assess the quality of test suites. Despite being used frequently, mutation testing is not so commonly applied in the mobile world. One critical challenge in mutation testing is dealing with its computational cost. Generating mutants, running test cases over each mutant, and analyzing the results may require significant time and resources. This research aims to contribute to reducing Android mutation testing costs. It implements mutation testing operators (traditional and Android-specific) according to mutant schemata (implementing multiple mutants into a single code file). It also describes an Android mutation testing framework developed to execute test cases and determine mutation scores. Additional mutation operators can be implemented in JavaScript and easily integrated into the framework. The overall approach is validated through case studies showing that mutant schemata have advantages over the traditional mutation strategy (one file per mutant). The results show mutant schemata overcome traditional mutation in all evaluated aspects with no additional cost: it takes 8.50% less time for mutant generation, requires 99.78% less disk space, and runs, on average, 6.45% faster than traditional mutation. Moreover, considering sustainability metrics, mutant schemata have 8,18% less carbon footprint than traditional strategy.

Abstract
The rapid integration of Machine Learning (ML) in organizational practices has driven demand for substantial computational resources, incurring both high economic costs and environmental impact, particularly from energy consumption. This challenge is amplified in dynamic data environments, where ML models must be frequently retrained to adapt to evolving data patterns. To address this, more sustainable Machine Learning Operations (MLOps) pipelines are needed for reducing environmental impacts while maintaining model accuracy. In this paper, we propose a model reuse approach based on data similarity metrics, which allows organizations to leverage previously trained models where applicable. We introduce a tailored set of meta-features to characterize data windows, enabling efficient similarity assessment between historical and new data. The effectiveness of the proposed method is validated across multiple ML tasks using the cosine and Bray-Curtis distance functions, which evaluate both model reuse rates and the performance of reused models relative to newly trained alternatives. The results indicate that the proposed approach can reduce the frequency of model retraining by up to 70% to 90% while maintaining or even improving predictive performance, contributing to more resource-efficient and sustainable MLOps practices.

Abstract
This paper proposes a multi-objective stochastic optimization framework that can be used by governments to run auctions and select the best community energy storage system (CESS) projects to support. The framework enables CESS providers and energy community members to equitably benefit from the economic value generated by CESSs. The auction accepts offers from competing CESS providers that constitute the data of the CESS location, size, install time, technology, provider, investment cost, and energy trading price. The auction is run by a government agency which selects CESS projects that maximize the economic benefits and distribute them equitably among CESS providers and community members. The multi-objective stochastic optimization accounts for the multi-year uncertainties of photovoltaic (PV) generation, real and reactive energy consumption, energy trading prices, and PV installations. We exploit the Monte Carlo simulation and scenario trees to model the aforementioned uncertainties. The K-Means clustering method is used to reduce the number of scenarios, and thereby, lessen the computational burden of the optimization problem. Our experiments on an Australian low-voltage network with a community of prosumers and consumers demonstrate that government financial support can accelerate the installation of CESSs and enhance their business viability. This can be achieved by boosting the economic benefits shared between CESS providers and communities and ensuring these benefits are distributed equitably. Also, our experiments show that the economic benefits of all stakeholders are further improved with a high growth of the number of PV installations, and a slight reduction of energy import and export prices over the planning period.

Abstract