Results 71 to 80 of about 20,293,735 (305)
A Green AI Methodology Based on Persistent Homology for Compressing BERT
Applied SciencesLarge Language Models (LLMs) like BERT have gained significant prominence due to their remarkable performance in various natural language processing tasks. However, they come with substantial computational and memory costs.
Luis Balderas +2 more
doaj +1 more source
Eddies: Fluid Dynamical Niches or Transporters?–A Case Study in the Western Baltic Sea
Frontiers in Marine Science, 2019 Fluid flows in the ocean have a strong impact on the growth and distribution of planktonic communities. In this case study, we applied a Lagrangian eddy detection and tracking tool and a transfer operator approach to data from a coupled hydrodynamical ...
Rahel Vortmeyer-Kley +5 more
doaj +1 more source
A combined robust fuzzy time series method for prediction of time series [PDF]
Neurocomputing, 2017 Outlier(s) have an adverse impact on the performance of fuzzy time series models.We proposed a combined robust fuzzy time series model (C-R-FTSM).C-R-FTSM uses fuzzy inputs composed of membership values as well as the crisp data.Training process of C-R-FTSM is performed by PSO in a single optimization process.Huber's loss function based on M estimator ...
Ozge Cagcag Yolcu, Hak-Keung Lam
openaire +6 more sources
Pediatric Blood &Cancer, EarlyView.ABSTRACT We assessed the effect of iron overload (IO) on mortality and complications following hematopoietic stem cell transplantation (HSCT) in patients with Diamond–Blackfan anemia syndrome (DBAS) in a systematic review of individual participant data and cohort data from observational studies.
Geoffrey Z. L. Kuppens +6 more
wiley +1 more source
Optimizing Convolutional Neural Network Architectures
MathematicsConvolutional neural networks (CNNs) are commonly employed for demanding applications, such as speech recognition, natural language processing, and computer vision.
Luis Balderas +2 more
doaj +1 more source
Neural Network Ensembles for Time Series Prediction [PDF]
, 2007 Rapidly evolving businesses generate massive amounts of time-stamped data sequences and defy a demand for massively multivariate time series analysis. For such data the predictive engine shifts from the historical auto-regression to modelling complex
Ruta, Dymitr +3 more
core +1 more source
Pediatric Blood &Cancer, EarlyView.ABSTRACT Purpose Next‐generation sequencing (NGS) has emerged as a promising approach to improve diagnostic accuracy, but its feasibility in low‐ and middle‐income countries remains unknown. This study characterized the diagnostic landscape and assessed organizational readiness for NGS implementation at two childhood cancer treatment centers in Accra ...
Melissa Carvalho +6 more
wiley +1 more source
Modelling Stabilometric Time Series [PDF]
, 2010 Stabilometry is a branch of medicine that studies balance-related human functions. Stabilometric systems generate time series. The analysis of these time series using data mining techniques can be very useful for domain experts.
Caraça-Valente Hernández, Juan Pedro +3 more
core
Time-series clustering via quasi U-statistics
, 2015 Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)The problem of time-series discrimination and classification is discussed. We propose a novel clustering algorithm based on a
Pinheiro, A, Valk, M
core +1 more source
Analyzing Multimodal Time Series as Dynamical Systems [PDF]
, 2010 We propose a novel approach to discovering latent structures from multimodal time series. We view a time series as observed data from an underlying dynamical system. In this way, analyzing multimodal time series can be viewed as finding latent structures
Chen Yu +3 more
core +1 more source