{"product_id":"9783032475633","title":"Irregularly Sampled Spatiotemporal Data A Graph Neural Network Perspective","description":"\u003ch3\u003eSpringerBriefs in Applied Sciences and Technology PoliMI SpringerBriefs\u003c\/h3\u003e\u003ch1\u003eIrregularly Sampled Spatiotemporal Data\u003c\/h1\u003e\u003ch2\u003eA Graph Neural Network Perspective\u003c\/h2\u003e\u003ch3\u003eCesare Alippi | Ivan Marisca\u003c\/h3\u003e\u003cdiv\u003e\u003cb\u003eComputers \/ Data Science \/ General\u003c\/b\u003e\u003c\/div\u003e\u003cbr\u003e\u003cdiv\u003e\n\u003cp\u003eThis book offers a comprehensive, graph-based deep learning framework for modeling and analyzing irregular spatiotemporal data, providing principled solutions where missing values, asynchronous sampling, and complex non Euclidean structures are the rule rather than the exception. It bridges the gap between theory and application to deliver state-of-the-art machine learning methodologies for reconstructing, filtering, and predicting time multivariate series across complex networks.\u003c\/p\u003e\n\u003cp\u003eModern data streams from traffic networks, environmental monitoring stations, smart energy grids, financial markets, and Internet-of-Things infrastructures rarely produce clean data. Observations are frequently missing due to sensor malfunctions, electronic readout or communication errors, sampled at irregular frequencies or misaligned over time. This book addresses these foundational challenges by treating spatiotemporal data as discrete signals evolving on graphs, where nodes represent data-generating entities and edges encode relationships among them.\u003c\/p\u003e\n\u003cp\u003eReaders will explore the intersection of graph signal processing and geometric deep learning, focusing on three core tasks: imputation, filtering, and prediction. The work goes beyond treating data irregularity and extreme sparsity as simple preprocessing obstacles; instead, it reframes them as inherent components of the learning architecture. By exploring message-passing mechanisms and advanced attention-based models, it explains how to design robust networks that maintain spatial and temporal consistency even under extreme data sparsity. Crucially, it tackles the scalability bottleneck, showing how to balance model expressiveness with computational efficiency when processing thousands of sensors over long temporal horizons.\u003c\/p\u003e\n\u003c\/div\u003e\u003cbr\u003e\u003ctable\u003e\n\u003ctr\u003e\n\u003ctd\u003ePublication Date: \u003c\/td\u003e\n\u003ctd\u003e28 February 2027\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003ePublisher: \u003c\/td\u003e\n\u003ctd\u003eSpringer Nature Switzerland\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eImprint: \u003c\/td\u003e\n\u003ctd\u003eSpringer\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eISBN-13: \u003c\/td\u003e\n\u003ctd\u003e9783032475633\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eFormat: \u003c\/td\u003e\n\u003ctd\u003ePaperback softback\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003c\/table\u003e","brand":"Springer Nature Switzerland","offers":[{"title":"Default Title","offer_id":63762275664012,"sku":"9783032475633","price":53.99,"currency_code":"USD","in_stock":true}],"url":"https:\/\/lateknightbooks.com\/products\/9783032475633","provider":"Late Knight Books and Services, LLC","version":"1.0","type":"link"}