Rare-class learning over Mg-doped ZnO nanoparticles

Kurban H., Kurban M.

CHEMICAL PHYSICS, cilt.546, 2021 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 546
  • Basım Tarihi: 2021
  • Doi Numarası: 10.1016/j.chemphys.2021.111159
  • Dergi Adı: CHEMICAL PHYSICS
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Chemical Abstracts Core, INSPEC
  • Anahtar Kelimeler: Machine learning, Material science, Rare-class learning, Tree-based models, Extreme gradient boosting, NEURAL-NETWORKS, DISCRIMINANT-ANALYSIS, PARAMETRIZATION, CLASSIFICATION, REGRESSION
  • Ankara Üniversitesi Adresli: Hayır

Özet

This interdisciplinary study is conducted to find answers to two important questions which researchers often face in Machine Learning (ML) and Material Science (MS) fields. In this work, we measure the performance of the most popular ML algorithms (more than a dozen) on rare-class learning problem and determine the best learning algorithm for atom type prediction over the Mg-doped ZnO nanoparticles data obtained from the densityfunctional tight-binding method. As a result, we observe that tree-based ML algorithms such as Extreme Gradient Boosting (XGB), Decision Trees (DT), Random Forest (RF), outperform other types of ML algorithms, e. g., cost-sensitive learning, prototype models, support vector machines, kernel methods, on both rare-class learning and atom type prediction.