Emine Kucukbenli's research aims to explore the vast landscape of crystal structures that atoms or molecules form. She builds numerical tools to speed up the exploration using machine learning [1,2] or to identify different points on this landscape that correspond to different material properties such as NMR or phonon spectrum [3-7]. Her work is mainly focused on numerical tool building rather than the applications themselves, but sometimes she stumbles upon mysteries of elusive crystals  or puzzling experimental findings  that draw her curiosity. She supports open source Density Functional Theory software packages through development and enhancement of numerical methods  and she is a passionate advocate for reproducibility in science [11,12].
Emine Kucukbenli holds a BSc in Physics from Bilkent University, Turkey and PhD in Numerical and Theoretical Condensed Matter Physics from SISSA, Italy. Prior to her appointment at Harvard University, she was a postdoctoral researcher for E-CAM Centre of Excellence at SISSA and at Theos research group at EPFL, Switzerland. She was generously supported by non-governmental organizations such as ICTP, TWAS, AUST, OWSD/GenderInSITE during her work for and with under-represented communities in science as organizer and participant of workshops, lecture series, conference panels and outreach programs.
R Lot, F Pellegrini, Y Shaidu, E Kucukbenli, "PANNA: Properties from Artificial Neural Network Architectures", in preparation (2019).
F Pellegrini, R Lot, Y Shaidu, E Kucukbenli, "Reproducibility and transfer of knowledge in deep neural networks for electronic structure" in preparation (2019).
DOI: 10.1021/acs.jpcc.8b05689 ,