دانلود رایگان مقاله لاتین شبکه عصبی برای مواد انفورماتیک از سایت الزویر
عنوان فارسی مقاله:
ابعاد یکسان داده با شبکه های عصبی برای مواد انفورماتیک
عنوان انگلیسی مقاله:
Uniforming the dimensionality of data with neural networks for materials informatics
سال انتشار : 2016
برای دانلود رایگان مقاله شبکه عصبی برای مواد انفورماتیک اینجا کلیک نمایید.
مقدمه انگلیسی مقاله:
1. Introduction
The development ofmaterials informatics has resulted in signifi- cant progress in the modeling and prediction of material properties, thereby reducing the costs of real-world experiments, and is becoming a promising research field for soft computing (for example, [2–4]). By using soft computing techniques such as neural networks, evolutionary and genetic algorithms, and fuzzy modeling, materials scientists can more effectively search for novel materials. These techniques are used alone and in combination with quantum calculations for materials design. For example, a method combining density functional theory and an evolutionary algorithm was used to predict the crystal structure of LiBeH3 (Hu et al. [5]). By organizing the data into a material database, researchers can determine the relationships between material properties (for example, conductivity, the critical temperature of superconductors, and melting temperature) and the properties (for example, atomic number, atomic mass, and electron negativity) of the elements in the material. The properties of elements or their combinations are referred to as descriptors (or “features” in computer science). Once the relevant features are obtained, the predictions of properties and the modeling of materials becomes easier. The literature [6] describes five descriptor categories: constitutional, topological, physicochemical, structural, and quantum-chemical. For instance, Seko et al. [7] adopted the sum and product of the element properties, such as atomic number, atomic mass, and number of valence electrons as features involved in the prediction of the melting temperature of single and binary compounds. These are constitutional descriptors. In addition, the use of sum and product operations is based on the domain knowledge of the compounds, and is also found in [8].
برای دانلود رایگان مقاله شبکه عصبی برای مواد انفورماتیک اینجا کلیک نمایید.
کلمات کلیدی:
[PDF]A data analytic methodology for materials informatics - CiteSeerX citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.636.5348&rep=rep1... by OY Abuomar - 2014 - Related articles May 16, 2014 - A data analytic materials informatics methodology is proposed after applying ... artificial neural network (ANN) model was used to explore the ... Materials Informatics - Rajan - 2009 - Statistical Analysis and Data ... onlinelibrary.wiley.com/doi/10.1002/sam.10022/abstract by K Rajan - 2009 - Cited by 1 - Related articles Mar 19, 2009 - Statistical Analysis and Data Mining: The ASA Data Science Journal. Explore this journal >. Statistical Analysis and ... Materials Informatics ... Perspective: Materials informatics and big data: Realization of the ... aip.scitation.org/doi/full/10.1063/1.4946894 by A Agrawal - 2016 - Cited by 16 - Related articles The need for data informatics is also emphasized by the Materials Genome ...... In particular, neural networks, decision trees, and multivariate polynomial ... RESEARCH ASSISTANT IN MATERIALS INFORMATICS (IWM-2016-52) https://recruiting.fraunhofer.de/Vacancies/28957/Description/2 The position includes the development of novel big-data and machine learning paradigms, including artificial neural networks to better predict materials ... Advances in Computer Science and Information Technology: First ... https://books.google.com/books?isbn=364217857X Natarajan Meghanathan, B.K. Kaushik, Dhinaharan Nagamalai - 2010 - Computers rich for materials informatics[14],[15],[16],[17]. ... These rules are applied to somewhat limited materials data sets of ... Data mining involves some high-effective computational algorithms[18],[19], such as neural networks, genetic algorithm, etc.
