Neural networks applied in control loops and safety-critical domains have to meet more requirements than just the overall best function approximation. On the one hand, a small approximation error is required; on the other hand, the smoothness and the monotonicity of selected input-output relations have to be guaranteed. Otherwise, the stability of most of the control laws is lost. In this article we compare two neural network-based approaches incorporating partial monotonicity by structure, namely the Monotonic Multi-Layer Perceptron (MONMLP) network and the Monotonic MIN-MAX (MONMM) network. We show the universal approximation capabilities of both types of network for partially monotone functions. On a number of datasets, we investigate the advantages and disadvantages of these approaches related to approximation performance, training of the model and convergence.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1016/j.neunet.2009.09.002 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Enhanced Mechanical Properties in Bulk Nanograined Ni with High-Density Fivefold Twins.
Small
January 2025
Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, Shenyang, 110016, China.
Fivefold twins are extensively present in nanoparticles and nanowires, enhancing their performance in physical, chemical, and mechanical properties. However, a deep insight into the correlation between mechanical properties and fivefold twins in bulk nanograined materials is lacking due to synthesis difficulties. Here, a bulk fivefold-twinned nanograined Ni is synthesized via electrodeposition.
View Article and Find Full Text PDFEngineering a high-throughput clone for industrial-scale production of long-acting GLP-1 analogue with retained bio-efficacy.
Biotechnol Prog
January 2025
Department of Biotechnology, School of Bioengineering, SRM Institute of Science and Technology, Kattankulathur, Tamil Nadu, India.
Type 2 diabetes mellitus (T2DM) and obesity are critical global health issues with rising incidence rates. Glucagon-like peptide-1 (GLP-1) analogues have emerged as effective treatments due to their ability to regulate blood glucose levels and gastric emptying through central nervous signals involving hypothalamic receptors, such as leptin. To address the short plasma half-life of native GLP-1, a C-16 fatty acid was conjugated to lysine in the GLP-1 analogue sequence to enhance its longevity.
View Article and Find Full Text PDFKalmanFormer: using transformer to model the Kalman Gain in Kalman Filters.
Front Neurorobot
January 2025
Software Engineering College, Zhengzhou University of Light Industry, Zhengzhou, China.
Introduction: Tracking the hidden states of dynamic systems is a fundamental task in signal processing. Recursive Kalman Filters (KF) are widely regarded as an efficient solution for linear and Gaussian systems, offering low computational complexity. However, real-world applications often involve non-linear dynamics, making it challenging for traditional Kalman Filters to achieve accurate state estimation.
View Article and Find Full Text PDFIntegrating deep learning in public health: a novel approach to PICC-RVT risk assessment.
Front Public Health
January 2025
Beijing Hospital, National Center of Gerontology, Institute of Geriatric Medicine, Chinese Academy of Medical Sciences, Beijing, China.
Background: Machine learning is pivotal for predicting Peripherally Inserted Central Catheter-related venous thrombosis (PICC-RVT) risk, facilitating early diagnosis and proactive treatment. Existing models often assess PICC-RVT risk as static and discrete outcomes, which may limit their practical application.
Objectives: This study aims to evaluate the effectiveness of seven diverse machine learning algorithms, including three deep learning and four traditional machine learning models, that incorporate time-series data to assess PICC-RVT risk.
Genotype-function-phenotype correlations for SCN1A variants identified by clinical genetic testing.
Ann Clin Transl Neurol
January 2025
Department of Pharmacology, Northwestern University Feinberg School of Medicine, Chicago, Illinois, USA.
Objective: Interpretation of clinical genetic testing, which identifies a potential genetic etiology in 25% of children with epilepsy, is limited by variants of uncertain significance. Understanding functional consequences of variants can help distinguish pathogenic from benign alleles. We combined automated patch clamp recording with neurophysiological simulations to discern genotype-function-phenotype correlations in a real-world cohort of children with SCN1A-associated epilepsy.
View Article and Find Full Text PDFWant 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!