14 papers · ranked by Valyu relevance
Qian Li, Sanyang Liu, Xin‐She Yang
All metaheuristic optimization algorithms require some initialization, and the initialization for such optimizers is usually carried out randomly. However, initialization can have some significant influence on the performance of such algorithms. This paper presents a systematic comparison of 22 different initialization…
Takashi Katoh, Keisuke Goto
Initializing all elements of an array to a specified value is a basic operation that frequently appears in numerous algorithms and programs. Initializable arrays are abstract arrays that support initialization as well as reading and writing of any element of the array in less than linear time proportional to the length…
S. Westerhof, T. Hofman
This paper presents an optimization framework for Spatial Packaging of Interconnected Systems with Physical Interactions (SPI2) that addresses the geometric challenges of three-dimensional component placement and routing. While SPI2 generally includes physical interactions, this study isolates the spatial optimization…
Oluwatumininu Emmanuel Ayo-Ojo, Akpevweoghene Ogheneowho Ugono, Nkosinathi Dlamini
Accurate initialization of polymer architectures remains a critical yet underappreciated determinant of reliability in molecular dynamics simulations of soft matter systems. Errors in coordinate generation and connectivity assignment frequently introduce artificial stresses, topological inconsistencies, and numerical…
Authors not listed
Bayesian Optimization (BO) has become a standard tool for hyperparameter tuning in machine learning due to its sample efficiency when exploring expensive black-box functions. While most BO pipelines begin with uniform random initialization, default hyperparameter values shipped with popular ML libraries such as…
Georgios C. Chasparis, Michael Rossbory, Verena Haunschmid
We introduce an evolutionary stochastic-local-search (SLS) algorithm for addressing a generalized version of the so-called 1/V/D/R cutting-stock problem. Cutting-stock problems are encountered often in industrial environments and the ability to address them efficiently usually results in large economic benefits.…
Mengyu Huang, Yuxing Zhong, Huiwen Yang, Jiazheng Wang + 3 more
'Bo Bai' 'Ling Shi'] The simplex method is one of the most fundamental technologies for solving linear programming (LP) problems and has been widely applied to different practical applications. In the past literature, how to improve and accelerate the simplex method has attracted plenty of research. One important way…
Subhasish Basak, Sébastien Petit, Julien Bect, Emmanuel Vázquez
This article investigates the origin of numerical issues in maximum likelihood parameter estimation for Gaussian process (GP) interpolation and investigates simple but effective strategies for improving commonly used opensource software implementations. This work targets a basic problem but a host of studies…
Alejandro Mata Ali, Iñigo Perez Delgado, Marina Ristol Roura, Aitor Moreno-Fernández-de-Leceta
'Aitor Moreno-Fernández-de-Leceta'] We present a novel method for initializing layers of tensorized neural networks in a way that avoids the explosion of the parameters of the matrix it emulates. The method is intended for layers with a high number of nodes in which there is a connection to the input or output of all…
Tao Chen, Jingtong Zhang, Jing Liu, Youjin Deng + 1 more
Tao Chen,1, 2, ∗ Jingtong Zhang,3, 4, ∗ Jing Liu,3, † Youjin Deng,1, 2, ‡ and Pan Zhang3, § 1 Hefei National Laboratory for Physical Sciences at the Microscale and Department of Modern Physics, University of Science and Technology of China, Hefei 230026, China 2 Hefei National Laboratory, University of Science and…
Sheheryar Zaidi, Tudor Berariu, Hyunjik Kim, Jörg Bornschein + 3 more
'Claudia Clopath' 'Yee Whye Teh' 'Razvan Pascanu'] Re-initializing a neural network during training has been observed to improve generalization in recent works. Yet it is neither widely adopted in deep learning practice nor is it often used in state-of-the-art training protocols. This raises the question of when…
Armen Aghajanyan
Initialization of parameters in deep neural networks has been shown to have a big impact on the performance of the networks (Mishkin & Matas, 2015). The initialization scheme devised by He et al, allowed convolution activations to carry a constrained mean which allowed deep networks to be trained effectively (He et…
Xinyu Zhang, Mário Antunes, Tyler Estro, Erez Zadok + 1 more
Initialization profoundly affects evolutionary algorithm (EA) efficacy by dictating search trajectories and convergence. This study introduces a hybrid initialization strategy combining empty-space search algorithm (ESA) and opposition-based learning (OBL). OBL initially generates a diverse population, subsequently…
Kemelli C. Estacio-Hiroms, Ernesto E. Prudencio, Nicholas Malaya, Manav Vohra + 1 more
'Manav Vohra' 'Damon McDougall'] Copyright © 2008-2013 The PECOS Development Team, http://pecos.ices.utexas.edu Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation; with…