Informatics for Materials Science and Engineering

Informatics for Materials Science and Engineering

Data-driven Discovery for Accelerated Experimentation and Application
2013, Pages 71-95
Informatics for Materials Science and Engineering

Chapter 5 - Evolutionary Data-Driven Modeling

https://doi.org/10.1016/B978-0-12-394399-6.00005-9Get rights and content

Artificial neural networks (ANNs) and genetic programming (GP) have already emerged as two very effective computing strategies for constructing data-driven models for systems of scientific and engineering interest. However, coming up with accurate models or meta-models from noisy real-life data is often a formidable task due to their frequent association with high degrees of random noise, which might render an ANN or GP model either over- or underfitted. This problem has recently been tackled in two emerging algorithms, Evolutionary Neural Net (EvoNN) and Bi-objective Genetic Programming (BioGP), which utilize the concept of Pareto tradeoff and apply a bi-objective genetic algorithm (GA) in the basic framework of both ANNs and GP. These concepts are elaborated in detail in this chapter.

References (0)

Cited by (20)

  • Hybrid optimization approach using evolutionary neural network & genetic algorithm in a real-world waterflood development

    2022, Journal of Petroleum Science and Engineering
    Citation Excerpt :

    The bi-objective genetic algorithm is applied to carry out a tradeoff between the error of the network training and the complexity of the network. The prey populations surviving after generations of predator attack will then represent the optimum Pareto solutions (Chakraborti, 2013; Li, 2003; Mondal et al., 2011; Pettersson et al., 2007, 2009). EvoNN algorithm was used extensively in materials and manufacturing processes such as blast furnace analyses and optimization of mechanical properties of micro-alloyed steel (Chugh et al., 2017; Laumanns et al., 1998; Li et al., 2007).

  • Bi-objective optimal design of a damage-tolerant multifunctional battery system

    2016, Materials and Design
    Citation Excerpt :

    The Latin hypercube sampling (LHS) method [33] is adopted to achieve uniformity and balance of sampling points in the design window. The metamodels are constructed from the FE results at sampling points using an evolutionary neural network (EvoNN) approach [23,25,27]. Then, after being validated and tested, the metamodels are used to approximate quantities of interest.

  • Hybrid Multi-Objective Optimization Approach in Water Flooding

    2022, Journal of Energy Resources Technology, Transactions of the ASME
View all citing articles on Scopus
View full text