This Special Issue consists of articles on research presented in the Second International Conference on Data-Driven Plasma Science (ICDDPS-2), which was held in Marseille, France, May 13–17, 2019.1 The ICDDPS-2 follows the initiative launched in 2018 by Sadruddin Benkadda, Deborah O'Connell, and Satoshi Hamaguchi to establish a regular international conference series that gathers a wide spectrum of plasma scientists interested in emerging data-driven methods applied to various topics in plasma science, including plasma-based future medical therapeutics such as cancer treatment, plasma sterilization, sophisticated materials synthesis, and coatings, advanced plasma processing for the manufacturing of next-generation semiconductor devices, and plasma-based clean energy technologies such as thermonuclear fusion and other non-fossil energy sources. This conference series aims to stimulate dialog and collaboration among researchers from the wide disciplines of plasma and data sciences.
Although data science in plasma science and engineering has a long history dating back to the early 1990s, when the first neural-network analyses were applied to semiconductor manufacturing processes, including plasma etching2–4 as well as the control of disruptions in tokamaks,5,6 the need to build a community of plasma scientists aware of the importance of new developments and challenges in data-driven science has become acute only recently. Indeed, advances in experimental and computational plasma research are producing larger and more complex data sets. Deriving values from the increasing volume and diversity of data now available to the research community provides opportunities for greater insight into complex plasma systems and their interfaces with materials and other phases. To maximize meaningful values obtainable from such data and fully facilitate the interpretation of processed data, the development of new techniques, possibly specific to plasma science, is now urgently required. Embracing data-driven discovery in plasma science will enable a new era of research and development in plasma science and technology.
The papers of ICDDPS-2 published in the current special issue7–13 represent a small part of the ongoing research worldwide in this field, in which the number of published papers has been exponentially growing for the last five years. Many novel algorithms and analytic techniques specifically designed to solve problems in plasma science have been developed. We hope such techniques and methodologies will lead to an ultimate understanding of highly complex nonlinear phenomena typically encountered in the various branches of plasma science.
The Guest Editors express thanks to the Physics of Plasmas editorial office staff for the encouragement and assistance in the production of this special issue. They would also like to thank the numerous colleagues who acted as anonymous referees for the articles published in this issue. The ICDDPS-2 and this Special Topic were partially supported by Aix Marseille University, CNRS, and JSPS Core-to-Core Program JPJSCCA2019002.