Wepresent an efficient implementation of the Zero Static Internal Stress Approximation(ZSISA)within the Quasi-Harmonic Approximation framework to compute anisotropic thermal expansion and elastic constants from first ...Wepresent an efficient implementation of the Zero Static Internal Stress Approximation(ZSISA)within the Quasi-Harmonic Approximation framework to compute anisotropic thermal expansion and elastic constants from first principles.By replacing the costly multidimensional minimization with a gradientbased method that leverages second-order derivatives of the vibrational free energy,the number of required phonon band structure calculations is significantly reduced:only six are needed for hexagonal,trigonal,and tetragonal systems,and 10–28 for lower-symmetry systems to determine the temperature dependence of lattice parameters and thermal expansion.This approach enables accurate modeling of anisotropic thermal expansion while substantially lowering computational cost compared to standard ZSISA method.The implementation is validated on a range of materials with symmetries from cubic to triclinic and is extended to compute temperature-dependent elastic constants with only a few additional phonon band structure calculations.展开更多
In this paper, the constrained optimization technique for a substantial prob-lem is explored, that is accelerating training the globally recurrent neural net-work. Unlike most of the previous methods in feedforward ne...In this paper, the constrained optimization technique for a substantial prob-lem is explored, that is accelerating training the globally recurrent neural net-work. Unlike most of the previous methods in feedforward neuxal networks, the authors adopt the constrained optimization technique to improve the gradiellt-based algorithm of the globally recuxrent neural network for the adaptive learn-ing rate during training. Using the recurrent network with the improved algo-rithm, some experiments in two real-world problems, namely filtering additive noises in acoustic data and classification of temporal signals for speaker identifi-cation, have been performed. The experimental results show that the recurrent neural network with the improved learning algorithm yields significantly faster training and achieves the satisfactory performance.展开更多
基金supported by the Fonds de la Recherche Scientifique (FRS-FNRS, Belgium) through the PdR Grant no. T.0103.19 – ALPS. It is an outcome of the Shapeable 2D magnetoelectronics by design project (SHAPEme, EOS Project No. 560400077525) that has received funding from the FWO and FRS-FNRS under the Belgian Excellence of Science (EOS) programComputational resources have been provided by the supercomputing facilities of the Université catholique de Louvain (CISM/UCL) and the Consortium des Equipements de Calcul Intensif en Fédération Wallonie Bruxelles (CECI), funded by the FRS-FNRS under Grant no. 2.5020.11.
文摘Wepresent an efficient implementation of the Zero Static Internal Stress Approximation(ZSISA)within the Quasi-Harmonic Approximation framework to compute anisotropic thermal expansion and elastic constants from first principles.By replacing the costly multidimensional minimization with a gradientbased method that leverages second-order derivatives of the vibrational free energy,the number of required phonon band structure calculations is significantly reduced:only six are needed for hexagonal,trigonal,and tetragonal systems,and 10–28 for lower-symmetry systems to determine the temperature dependence of lattice parameters and thermal expansion.This approach enables accurate modeling of anisotropic thermal expansion while substantially lowering computational cost compared to standard ZSISA method.The implementation is validated on a range of materials with symmetries from cubic to triclinic and is extended to compute temperature-dependent elastic constants with only a few additional phonon band structure calculations.
文摘In this paper, the constrained optimization technique for a substantial prob-lem is explored, that is accelerating training the globally recurrent neural net-work. Unlike most of the previous methods in feedforward neuxal networks, the authors adopt the constrained optimization technique to improve the gradiellt-based algorithm of the globally recuxrent neural network for the adaptive learn-ing rate during training. Using the recurrent network with the improved algo-rithm, some experiments in two real-world problems, namely filtering additive noises in acoustic data and classification of temporal signals for speaker identifi-cation, have been performed. The experimental results show that the recurrent neural network with the improved learning algorithm yields significantly faster training and achieves the satisfactory performance.
