Neural-Network-Biased Genetic Algorithms for Materials Design: Evolutionary Algorithms That Learn

被引:88
作者
Patra, Tarak K. [1 ]
Meenakshisundaram, Venkatesh [1 ]
Hung, Jui-Hsiang [1 ]
Simmons, David S. [1 ]
机构
[1] Univ Akron, Dept Polymer Engn, 250 South Forge St, Akron, OH 44325 USA
关键词
materials design; machine learning; optimization; neural network; genetic algorithm; Ising model; compatibilizer; polymers; soft matter; molecular dynamics simulation; GLASS-TRANSITION TEMPERATURE; PACKING DENSITY; POLYMER BLENDS; MOLECULAR-DYNAMICS; PREDICTION; OPTIMIZATION; COPOLYMERS; COMBINATORIAL; COMPATIBILIZERS; PERMEABILITY;
D O I
10.1021/acscombsci.6b00136
中图分类号
O69 [应用化学];
学科分类号
081704 ;
摘要
Machine learning has the potential to dramatically accelerate high-throughput approaches to materials design, as demonstrated by successes in biomolecular design and hard materials design. However, in the search for new soft materials exhibiting properties and performance beyond those previously achieved, machine learning approaches are frequently limited by two shortcomings. First, because they are intrinsically interpolative, they are better suited to the optimization of properties within the known range of accessible behavior than to the discovery of new materials with extremal behavior. Second, they require large pre-existing data sets, which are frequently unavailable and prohibitively expensive to produce. Here we describe a new strategy, the neural network-biased genetic algorithm (NBGA), for combining genetic algorithms, machine learning, and high-throughput computation or experiment to discover materials with extremal properties in the absence of pre-existing data. Within this strategy, predictions from a progressively constructed artificial neural network are employed to bias the evolution of a genetic algorithm, with fitness evaluations performed via direct simulation or experiment. In effect, this strategy gives the evolutionary algorithm the ability to "learn" and draw inferences from its experience to accelerate the evolutionary process. We test this algorithm against several standard optimization problems and polymer design problems and demonstrate that it matches and typically exceeds the efficiency and reproducibility of standard approaches including a direct-evaluation genetic algorithm and a neural-network evaluated genetic algorithm. The success of this algorithm in a range of test problems indicates that the NBGA provides a robust strategy for employing informatics-accelerated high-throughput methods to accelerate materials design in the absence of preexisting data.
引用
收藏
页码:96 / 107
页数:12
相关论文
共 89 条
  • [1] Prediction of high weight polymers glass transition temperature using RBF neural networks
    Afantitis, A
    Melagraki, G
    Makridima, K
    Alexandridis, A
    Sarimveis, H
    Iglessi-Markopoulou, O
    [J]. JOURNAL OF MOLECULAR STRUCTURE-THEOCHEM, 2005, 716 (1-3): : 193 - 198
  • [2] Basic concepts of artificial neural network (ANN) modeling and its application in pharmaceutical research
    Agatonovic-Kustrin, S
    Beresford, R
    [J]. JOURNAL OF PHARMACEUTICAL AND BIOMEDICAL ANALYSIS, 2000, 22 (05) : 717 - 727
  • [3] Amis EJ, 2004, NAT MATER, V3, P83, DOI 10.1038/nmat1064
  • [4] Anderson C. A., 1991, Complex Systems, V5, P327
  • [5] Nanostructured materials for advanced energy conversion and storage devices
    Aricò, AS
    Bruce, P
    Scrosati, B
    Tarascon, JM
    Van Schalkwijk, W
    [J]. NATURE MATERIALS, 2005, 4 (05) : 366 - 377
  • [6] Molecular designing of novel ternary copolymers of donor-acceptor polymers using genetic algorithm
    Arora, Vinita
    Bakhshi, A. K.
    [J]. CHEMICAL PHYSICS, 2010, 373 (03) : 307 - 312
  • [7] Adaptive Strategies for Materials Design using Uncertainties
    Balachandran, Prasanna V.
    Xue, Dezhen
    Theiler, James
    Hogden, John
    Lookman, Turab
    [J]. SCIENTIFIC REPORTS, 2016, 6
  • [8] Prediction and optimization of mechanical properties of polypropylene/waste tire powder blends using a hybrid Artificial Neural Network-Genetic Algorithm (GA-ANN)
    Balasubramanian, Maridass
    Paglicawan, Marissa A.
    Zhang, Zhen-Xiu
    Lee, Sung Hyo
    Xin, Zhen-Xiang
    Kim, Jin Kuk
    [J]. JOURNAL OF THERMOPLASTIC COMPOSITE MATERIALS, 2008, 21 (01) : 51 - 69
  • [9] Baluja S., 1995, MACHINE LEARNING
  • [10] Random-close packing limits for monodisperse and polydisperse hard spheres
    Baranau, Vasili
    Tallarek, Ulrich
    [J]. SOFT MATTER, 2014, 10 (21) : 3826 - 3841