Results 51 to 60 of about 57,862 (263)

Novel folded-PCA for improved feature extraction and data reduction with hyperspectral imaging and SAR in remote sensing

, 2014
As a widely used approach for feature extraction and data reduction, Principal Components Analysis (PCA) suffers from high computational cost, large memory requirement and low efficacy in dealing with large dimensional datasets such as Hyperspectral ...
Han, Junwei, Zabalza, Jaime, Yang, Mingqiang, Zhang, Yi, Ren, Jinchang, Marshall, Stephen, Wang, Jun   +6 more
core   +1 more source

M6AMRFS: Robust Prediction of N6-Methyladenosine Sites With Sequence-Based Features in Multiple Species

Frontiers in Genetics, 2018
As one of the well-studied RNA methylation modifications, N6-methyladenosine (m6A) plays important roles in various biological progresses, such as RNA splicing and degradation, etc.
Xiaoli Qiang, Huangrong Chen, Xiucai Ye, Ran Su, Leyi Wei   +4 more
doaj   +1 more source

Hospitalization Through Families’ Eyes: Comparing Inpatient Care Quality for Children With Sickle Cell Disease and Cystic Fibrosis in Canada

Pediatric Blood &Cancer, EarlyView.
ABSTRACT Background Sickle cell disease (SCD) is a chronic, inherited hemoglobinopathy that requires frequent hospitalization for disease‐related complications. Canadian data on inpatient care is limited. This study compared caregiver‐reported hospital experiences of children with SCD to those with cystic fibrosis (CF), a chronic, autosomal recessive ...
Hailey M. Zwicker, Kyle Kemp, Perri Tutelman, Sharon H. J. Hou, Paul Fairie, Maria Santana, Gregory M. T. Guilcher, Taryn Fay‐McClymont, Glenda Bendiak, Wendy Pelletier, Nicola Wright, Fiona Schulte   +11 more
wiley   +1 more source

Inverse Feature Learning: Feature Learning Based on Representation Learning of Error

IEEE Access, 2020
This paper proposes inverse feature learning (IFL) as a novel supervised feature learning technique that learns a set of high-level features for classification based on an error representation approach. The key contribution of this method is to learn the
Behzad Ghazanfari, Fatemeh Afghah, Mohammadtaghi Hajiaghayi   +2 more
doaj   +1 more source

Multi‐level cross‐modality learning framework for text‐based person re‐identification

Electronics Letters, 2023
The target of text‐based person re‐identification (Re‐ID) is to retrieve the corresponding image of a person through the given text information. However, due to the homogeneous variety and modality heterogeneity, it is challenging to simultaneously learn
Tinghui Wu, Shuhe Zhang, Dihu Chen, Haifeng Hu   +3 more
doaj   +1 more source

Contrastive-Learning-Based Time-Series Feature Representation for Parcel-Based Crop Mapping Using Incomplete Sentinel-2 Image Sequences

Remote Sensing, 2023
Parcel-based crop classification using multi-temporal satellite optical images plays a vital role in precision agriculture. However, optical image sequences may be incomplete due to the occlusion of clouds and shadows.
Ya’nan Zhou, Yan Wang, Na’na Yan, Li Feng, Yuehong Chen, Tianjun Wu, Jianwei Gao, Xiwang Zhang, Weiwei Zhu   +8 more
doaj   +1 more source

Venous Thromboembolism in Pediatric Bone Sarcoma Patients: A 10‐Year, Single‐Institution Experience Encompassing the COVID‐19 Pandemic

Pediatric Blood &Cancer, EarlyView.
ABSTRACT Background Osteosarcoma (OS) and Ewing sarcoma (EWS) are the most common primary bone cancers in children, but acute thrombosis is poorly characterized in this population. Our study evaluated the rates of venous thromboembolism (VTE) and associated risk factors in pediatric patients with bone sarcomas treated over a 10‐year period encompassing
Sarah Kappa, Shou Xu, Mark Zobeck, Sarah E. Sartain, Nino Rainusso, Lisa Wang, Nicole Montgomery, Paige Wheaton, Sarah Whittle   +8 more
wiley   +1 more source

Learning Hyperbolic Representations of Topological Features

CoRR, 2021
Learning task-specific representations of persistence diagrams is an important problem in topological data analysis and machine learning. However, current state of the art methods are restricted in terms of their expressivity as they are focused on Euclidean representations.
Panagiotis Kyriakis, Iordanis Fostiropoulos, Paul Bogdan   +2 more
openaire   +3 more sources

Deep Sequencing of FLT3‐ITD Enables Response Evaluation and Post‐Treatment Monitoring in Childhood AML: An Exploratory Study

Pediatric Blood &Cancer, EarlyView.
ABSTRACT Background An internal tandem duplication in the gene encoding Fms‐like tyrosine kinase 3 (FLT3‐ITD) is associated with high relapse risk and poor prognosis in acute myeloid leukemia (AML) and plays a crucial role in treatment decisions. Measurable residual disease (MRD) analysis of FLT3‐ITD during and after treatment has shown prognostic ...
Sofie Johansson Alm, Bea Tornberg, Erik Delsing Malmberg, Henrik Hasle, Kristian Løvvik Juul‐Dam, Lars Palmqvist, Borhan R. Saeed, Giti Shah Barkhordar, Anastasia Soboli, Anna Staffas, Jonas Abrahamsson, Linda Fogelstrand   +11 more
wiley   +1 more source

Geometric feature distributions for shape representation and recognition. [PDF]

, 1994
One of the fundamental problems in computer vision is the identification of objects from their shape. The research reported in this thesis is directed toward the development of a scheme for representing the shape of an object which allows it to be ...
Evans, Alun C
core