Classification of molecular structures is a crucial step in dynamics (MD) simulations to detect various and phases within systems. Molecular structures, which are commonly identified using order parameters, were recently machine learning (ML), that is, the ML models acquire structural features labeled crystals or via supervised learning. However, these approaches may not identify unlabeled unknown such as imperfect crystal observed nonequilibrium systems interfaces. In this study, we...

In this study, using some machine learning methods, we develop a framework that deals with forward analysis to predict property from polymer alloy's phase separation structure and inverse design generate the property. We only consider Young's modulus as in study. The is performed convolutional neural network (CNN) realized by random search toward model combining generative adversarial (GAN) CNN. This applicable other properties at low computational cost, latent variables belonging GAN are...

Clathrate hydrates continue to be the focus of active research efforts due their use in energy resources, transportation, and storage-related applications. Therefore, it is crucial define essential characteristics from a molecular standpoint. Understanding structure particular because aids understanding mechanisms that lead formation or dissociation clathrate hydrates. In past, wide variety order parameters have been employed classify evaluate hydrate structures. An alternative approach...

Solving continuous variable optimization problems by factorization machine quantum annealing (FMQA) demonstrates the potential of Ising machines to be extended as a solver for integer and real problems. However, details Hamiltonian function surface obtained (FM) have been overlooked. This study shows that in widely used case where numbers are represented combination binary variables, FM can very noisy. noise interferes with inherent capabilities is likely substantial cause previously...

ABSTRACT In this study, we developed a method of estimating the correction terms that makes Hamiltonian used in phase‐field analysis by quantum annealing correspond to free energy functional conventional using finite difference method. For estimation terms, employed factorization machine. The inputs machine were variables domain‐wall encoding and differences between Gibbs Hamiltonian. We obtained value quadratic unconstrained binary optimization (QUBO) form as output learning QUBO was...

Abstract Prediction of protein–ligand binding affinity is a major goal in drug discovery. Generally, free energy gap calculated between two states (e.g., ligand and unbinding). The implicitly includes the effects changes protein dynamics induced by binding. However, relationship remains unclear. Here, we propose method that represents ligand-binding-induced behavioral change with simple feature can be used to predict affinity. From unbiased molecular simulation data, an unsupervised deep...

Abstract Quasi-liquid layers (QLLs) are present on the surface of ice and play a significant role in its distinctive chemical physical properties. These exhibit considerable heterogeneity across different scales ranging from nanometers to millimeters. Although formation partially ice-like structures has been proposed, molecular-level understanding this remains unclear. Here, we examined molecular dynamics QLLs based simulations machine learning analysis simulation data. We demonstrated that...

Molecular dynamics (MD) is a powerful computational method for simulating molecular behavior. Deep neural networks provide novel of generating MD data efficiently, but there no architecture that mitigates the well-known exposure bias accumulated by multi-step generations. In this paper, we propose time series generator using deep network based on Wasserstein generative adversarial nets. Instead sparse real data, our model evolves latent variable z densely distributed in low-dimensional...

Pseudo-random number generators (PRNGs) are software algorithms generating a sequence of numbers approximating the properties random numbers. They critical components in many information systems that require unpredictable and nonarbitrary behaviors, such as parameter configuration machine learning, gaming, cryptography, simulation. A PRNG is commonly validated through statistical test suite, NIST SP 800-22rev1a (NIST suite), to evaluate its robustness randomness In this paper, we propose...

Molecular dynamics simulation produces three-dimensional data on molecular structures. The classification of structure is an important task. Conventionally, various order parameters are used to classify different structures liquid and crystal. Recently, machine learning (ML) methods have been proposed based find optimal choices or use them as input features neural networks. Conventional ML still require manual operation, such calculating the conventional manipulating impose...

A novel model to be applied next-generation accelerators, Ising machines, is formulated on the basis of phase-field phase-separation structure a diblock polymer. Recently, machines including quantum annealing attract overwhelming attention as technology that opens up future possibilities. On other hand, has demonstrated its high performance in material development, though it takes long time achieve equilibrium. Although convergence problem might solved by no solution been proposed. In this...

In this study, we developed a new method of topology optimization for truss structures by quantum annealing. To perform annealing analysis with real variables, representation numbers as sum random number combinations is employed. The nodal displacement expressed binary variables. Hamiltonian H formulated on the basis elastic strain energy and position structure. It confirmed that deformation possible For structure, cross-sectional area iterative calculation changes in leads to optimal...

Abstract This study examined the applicability of factorization machines with quantum annealing (FMQA) to field landslide risk assessment for two specific black‐box optimization problems, hyperparameter (HPO) metamodeling and metamodel‐based simulation (MBSO) targeting granular flow using discrete element method (DEM). These problems are solved successively: HPO is first performed determine hyperparameters Gaussian process regression (GPR) metamodel, which then used as a low‐cost, fast...

Molecular dynamics simulation is a method of investigating the behavior molecules, which useful for analyzing variety structural and dynamic properties mechanisms phenomena. However, huge computational cost large-scale long-time simulations an enduring problem that must be addressed. MD-GAN machine learning-based can evolve part system at any time step, accelerating generation molecular data [Endo et al., Proceedings AAAI Conference on Artificial Intelligence, 2018, 32]. For accurate...

Lubricants with desirable frictional properties are important in achieving an energy-saving society. at the interfaces of mechanical components confined under high shear rates and pressures behave quite differently from bulk material. Computational approaches such as nonequilibrium molecular dynamics (NEMD) simulations have been performed to probe behavior lubricants. However, low-shear-velocity regions materials rarely simulated owing expensive calculations necessary do so, velocities...

Self-consistent field (SCF) analysis is an indispensable tool for predicting the microphase separation structures of polymer alloys. However, computation phase-separated in equilibrium state computationally intensive, leading to high costs. To address this challenge, we propose a novel deep learning approach that leverages generative adversarial network (GAN), powerful model, accelerate SCF analysis. Specifically, trained GAN using comprehensive data obtained from analysis, enabling us...

Quantum annealing machines are next-generation computers for solving combinatorial optimization problems. Although physical simulations one of the most promising applications quantum machines, a method how to embed target problem into has not been developed except certain simple examples. In this study, we focus on representing real numbers using binary variables, or bits. One important problems conducting simulation by is represent number with The variables in often represented but must be...

Permeation through polymer membranes is an important technology in the chemical industry, and its design, self-diffusion coefficient one of physical quantities that determine permeability. Since sensitively reflects intra- intermolecular interactions, analysis using all-atom model required. However, simulations are computationally expensive require long simulation times for diffusion small molecules dissolved polymers. MD-GAN, a machine learning model, effective accelerating reducing...

Carbon-fiber-reinforced plastic (CFRP) is a composite material whose base and reinforcement carbon fibers. CFRP widely used in various fields for laminating prepregs. The laminated plate tends to sustain damage, such as delamination, fiber breakage, breakage; hence, we must conduct high-precision efficient nondestructive testing (NDT). Examples of NDT are ultrasonic examination, X-ray tomography, infrared stress analysis. With most methods, it difficult easily obtain detailed correct...

Generalized eigenvalue problems (GEPs) play an important role in the variety of fields including engineering, machine learning, and quantum chemistry. Especially, many these can be reduced to finding minimum or maximum GEPs. One key handle GEPs is that memory usage computational complexity explode as size system interest grows. This paper aims at extending sequential optimizers for Sequential are a family algorithms iteratively solve analytical optimization single-qubit gates coordinate...

Deep learning approach is applied to detect the representative molecular behavior from enormous and complex dynamics data.

We have incorporated Evolution Strategies into the Replica-Exchange Monte Carlo simulation method to predict phase behavior of several example fluids. The replica-exchange allows one system exchange temperatures with its neighbors search for most stable structure relatively efficiently in a single simulation. However, if temperature intervals replicas are not positioned carefully, there is an issue that local does occur. Our results simple Lennard-Jones fluid and liquid-crystal Yukawa model...

Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI

Solving continuous variable optimization problems by factorization machine quantum annealing (FMQA) demonstrates the potential of Ising machines to be extended as a solver for integer and real problems. However, details Hamiltonian function surface obtained (FM) have been overlooked. This study shows that in widely common case where numbers are represented combination binary variables, FM can very noisy. noise interferes with inherent capabilities is likely substantial cause previously...

In variational algorithms, quantum circuits are conventionally parametrized with respect to single-qubit gates. this study, we parameterize a generalized controlled gate and propose an algorithm estimate the optimal parameters for locally minimizing cost value, where extend free quaternion selection method, optimization method gate. To benchmark performance, apply proposed various problems, including Variational Quantum Eigensolver (VQE) Ising molecular Hamiltonians, Algorithms (VQA)...