Abstract:Structural characterization of nanoclusters is one of the major challenges in nanocluster modeling owing to the multitude of possible configurations of arrangement of cluster atoms. The genetic algorithm (GA), a class of evolutionary algorithms based on the principles of natural evolution, is a commonly employed search method for locating the global minimum configuration of nanoclusters. Although a GA search at the DFT level is required for the accurate description of a potential energy surface to arrive at the correct global minimum configuration of nanoclusters, computationally expensive DFT evaluation of the significantly larger number of cluster geometries limits its practicability. Recently, machine learning potentials (MLP) that are learned from DFT calculations gained significant attention as computationally cheap alternative options that provide DFT level accuracy. As the accuracy of the MLP predictions is dependent on the quality and quantity of the training DFT data, active learning (AL) strategies have gained significant momentum to bypass the need of large and representative training data. In this application note, we present Cluster-MLP, an on-the-fly active learning genetic algorithm framework that employs the Flare++ machine learning potential (MLP) for accelerating the GA search for global minima of pure and alloyed nanoclusters. We have used a modified version the Birmingham parallel genetic algorithm (BPGA) for the nanocluster GA search which is then incorporated into distributed evolutionary algorithms in Python (DEAP), an evolutionary computational framework for fast prototyping or technical experiments. We have shown that the incorporation of the AL framework in the BPGA significantly reduced the computationally expensive DFT calculations. Moreover, we have shown that both the AL-GA and DFT-GA predict the same global minima for all the clusters we tested.

Recent Progresses of Global Minimum Searches of Nanoclusters with a Constrained Basin-Hopping Algorithm in the TGMin Program

TGMin: an Efficient Global Minimum Searching Program for Free and Surface-Supported Clusters.

An approach based on genetic algorithms and machine learning coupled for studying alloy and molecular clusters by optimizing quantum energy surfaces

Unbiased Fuzzy Global Optimization of Lennard-Jones Clusters for N ≤ 1000.

Reliability of semiempirical and DFTB methods for the global optimization of the structures of nanoclusters

A flexible and adaptive grid algorithm for global optimization utilizing basin hopping Monte Carlo

PGA: A new particle swarm optimization algorithm based on genetic operators for the global optimization of clusters

Funnel Hopping: Searching the Cluster Potential Energy Surface over the Funnels.

Unraveling the Planar-Globular Transition in Gold Nanoclusters through Evolutionary Search

Parameter Optimization in Cluster Identification Algorithms for Characterizing Nanoclusters in Al–Mg–Si–Cu Alloys

Subnano Pt Particles from a First-Principles Stochastic Surface Walking Global Search.

Going for Gold(-Standard): Attaining Coupled Cluster Accuracy in Oxide-Supported Nanoclusters

Exploring Nanocluster Potential Energy Surfaces via Deep Reinforcement Learning: Strategies for Global Minimum Search

Global Optimization Method for Cluster Structures.

First principles global optimization of metal clusters and nanoalloys

Global Optimization of Molybdenum Subnanoclusters on Graphene: A Consistent Approach toward Catalytic Applications

Cluster-MLP: An Active Learning Genetic Algorithm Framework for Accelerated Discovery of Global Minimum Configurations of Pure and Alloyed Nanoclusters

An Improved Self-Adaptive Differential Evolution with the Neighborhood Search Algorithm for Global Optimization of Bimetallic Clusters

Probing the critical nucleus size in tetrahydrofuran clathrate hydrate formation using surface-anchored nanoparticles

Discovering Syntheses of Atomically Precise Metal Nanoclusters by Applying Machine Learning to a High‐Throughput Platform

Communication: Optimal Parameters for Basin-Hopping Global Optimization Based on Tsallis Statistics