arXiv subjects

Raul Zurita-Milla

Publications and source records attributed to Raul Zurita-Milla.

2 recordsLinked to original sources

arXiv2025

Energy Efficiency in Cloud-Based Big Data Processing for Earth Observation: Gap Analysis and Future Directions

Earth observation (EO) data volumes are rapidly increasing. While cloud computing are now used for processing large EO datasets, the energy efficiency aspects of such a processing have received much less attention. This issue is notable given the increasing awareness of energy costs and carbon footprint in big data processing, particularly with increased attention on compute-intensive foundation models. In this paper we identify gaps in energy efficiency practices within cloud-based EO big data (EOBD) processing and propose several research directions for improvement. We first examine the current EOBD landscape, focus on the requirements that necessitate cloud-based processing and analyze existing cloud-based EOBD solutions. We then investigate energy efficiency strategies that have been successfully employed in well-studied big data domains. Through this analysis, we identify several critical gaps in existing EOBD processing platforms, which primarily focus on data accessibility and computational feasibility, instead of energy efficiency. These gaps include insufficient energy monitoring mechanisms, lack of energy awareness in data management, inadequate implementation of energy-aware resource allocation and lack of energy efficiency criteria on task scheduling. Based on these findings, we propose the development of energy-aware performance monitoring and benchmarking frameworks, the use of optimization techniques for infrastructure orchestration, and of energy-efficient task scheduling approaches for distributed cloud-based EOBD processing frameworks. These proposed approaches aim to foster more energy awareness in EOBD processing , potentially reducing power consumption and environmental impact while maintaining or minimally impacting processing performance.

cs.DC↗

arXiv2025

On the use of adversarial validation for quantifying dissimilarity in geospatial machine learning prediction

Recent geospatial machine learning studies have shown that the results of model evaluation via cross-validation (CV) are strongly affected by the dissimilarity between the sample data and the prediction locations. In this paper, we propose a method to quantify such a dissimilarity in the interval 0 to 100% and from the perspective of the data feature space. The proposed method is based on adversarial validation, which is an approach that can check whether sample data and prediction locations can be separated with a binary classifier. The proposed method is called dissimilarity quantification by adversarial validation (DAV). To study the effectiveness and general?ity of DAV, we tested it on a series of experiments based on both synthetic and real datasets and with gradually increasing dissimilarities. Results show that DAV effectively quantified dissimilarity across the entire range of values. Next to this, we studied how dissimilarity affects CV methods' evaluations by comparing the results of random CV method (RDM-CV) and of two geospatial CV methods, namely, block and spatial+ CV (BLK-CV and SP-CV). Our results showed the evaluations follow similar patterns in all datasets and predictions: when dissimilarity is low (usually lower than 30%), RDM-CV provides the most accurate evaluation results. As dissimilarity increases, geospatial CV methods, especially SP-CV, become more and more accurate and even outperform RDM-CV. When dissimilarity is high (>=90%), no CV method provides accurate evaluations. These results show the importance of considering feature space dissimilarity when working with geospatial machine learning predictions and can help researchers and practitioners to select more suitable CV methods for evaluating their predictions.

cs.LG↗