The relationship between the number of hidden nodes in a neural network, the complexity of a multiclass discrimination problem, and the number of samples needed for effect learning are discussed. Bounds for the number of samples needed for effect learning are given. It is shown that Omega(min (d,n) M) boundary samples are required for successful classification of M clusters of samples using a two-hidden-layer neural network with d-dimensional inputs and n nodes in the first hidden layer.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1109/72.97932 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Comprehensive analysis of the interaction microbiome and prostate cancer: an initial exploration from multi-cohort metagenome and GWAS studies.
J Transl Med
January 2025
Department and Institute of Urology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, No.1095 Jiefang Avenue, Wuhan, Wuhan, 430030, P.R. China.
Introduction: Prostate cancer is one of the most common cancers in the United States with a high mortality rate. In recent years, the traditional opinion about prostate microbiome was challenged. Although there still are some arguments, an escalating number of researchers are shifting their focus toward the microbiome within the prostate tumor environment.
View Article and Find Full Text PDFA novel compound heterozygous mutation in the DYNC2H1 gene in a Chinese family with Jeune syndrome.
Hereditas
January 2025
Key Laboratory of Reproductive Health Diseases Research and Translation of Ministry of Education & Key Laboratory of Human Reproductive Medicine and Genetic Research of Hainan Provincie & Hainan Provincial Clinical Research Center for Thalassemia, The First Affiliated Hospital of Hainan Medical University, Hainan Medical University, Haikou, Hainan, 571101, China.
Background: The dynein cytoplasmic two heavy chain 1 (DYNC2H1) gene encodes a cytoplasmic dynein subunit. Cytoplasmic dyneins transport cargo towards the minus end of microtubules and are thus termed the "retrograde" cellular motor. Mutations in DYNC2H1 are the main causative mutations of short rib-thoracic dysplasia syndrome type III with or without polydactyly (SRTD3).
View Article and Find Full Text PDFSimulation-based training program effect on pediatric nurses' knowledge and performance regarding heel-prick during newborn blood screening test.
BMC Nurs
January 2025
Faculty of Applied Medical Sciences in Alnamas, University of Bisha, 255, Al Nakhil, Al-Namas, 67714, Saudi Arabia.
Background: One of the best ways to impart important skills to trainees is through simulation-based training, which is more reliable than the conventional clinical examination method. It is used in pediatric nursing training to allow nurses to practice and improve their clinical and conversational skills during an actual child encounter. A heel-prick is a complex psychomotor task that requires skill and knowledge from the pediatric nurse performing the procedure while applying for the National Newborn Screening Program.
View Article and Find Full Text PDFBridging past and present: exploring Cannabis traditions in Armenia through ethnobotanical interviews and bibliographic prospecting.
J Cannabis Res
January 2025
Laboratori de Botànica (UB), Facultat de Farmàcia i Ciències de l'Alimentació-Institut de Recerca de la Biodiversitat (IRBio), Unitat Associada al CSIC, Universitat de Barcelona, Av. Joan XXIII 27-31, Barcelona, Catalonia, 08028, Spain.
Background: Cannabis sativa L. (Cannabaceae) has been widely used by humans throughout its history for a variety of purposes (medicinal, alimentary and other uses). Armenia, with its rich cultural history and diverse ecosystems, offers a unique context for ethnobotanical research about traditional uses of Cannabis.
View Article and Find Full Text PDFTAD-SIE: sample size estimation for clinical randomized controlled trials using a Trend-Adaptive Design with a Synthetic-Intervention-Based Estimator.
Trials
January 2025
Department of Electrical and Computer Engineering, Princeton University, Princeton, 08544, NJ, USA.
Background: Phase-3 clinical trials provide the highest level of evidence on drug safety and effectiveness needed for market approval by implementing large randomized controlled trials (RCTs). However, 30-40% of these trials fail mainly because such studies have inadequate sample sizes, stemming from the inability to obtain accurate initial estimates of average treatment effect parameters.
Methods: To remove this obstacle from the drug development cycle, we present a new algorithm called Trend-Adaptive Design with a Synthetic-Intervention-Based Estimator (TAD-SIE) that powers a parallel-group trial, a standard RCT design, by leveraging a state-of-the-art hypothesis testing strategy and a novel trend-adaptive design (TAD).
Want AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!