An international group of scientists, including representatives of the Tampere University of Technology (Finland), University of Burgundy (France), Aston University (England), and NSU participated in a review of the potential for using machine learning methods in photon measurements and creating "smart" lasers in the future. An article presenting their work, “Machine Learning and Applications in Ultrafast Photonics”, was published in “Nature Photonics”, a prestigious international journal with an impact factor of 31.241.
Novosibirsk State University is a world leader in the ”smart” laser field. This is demonstrated by regular publications in leading international scientific journals. Sergey Kobtsev, Head of the NSU Laser Physics and Innovative Technologies Section and one of the review’s authors, talked about the work presented in the article,
Fiber lasers combine convenience in simple operations and inconvenience managing complex parameters. All of the work presented is focused on the search for tools and methods to control the generation parameters of short-pulse fiber lasers. Our work is another step in this direction. A remarkable property of these relatively new lasers is that fiber-optic technologies are organically combined with other modern technologies, for example, semiconductors. As a result of these combinations new quality is achieved.
Sergey Turitsyn, Head of the NSU Physics Department Laboratory of Nonlinear Photonics and another author added,
Fiber lasers are very interesting physical systems as well as applied engineering devices. The difficulty in controlling these lasers is caused by the nonlinear dynamics of light in the fiber optic cavity. On the one hand, this allows you to study fundamental nonlinear effects, and on the other, to apply the methods of nonlinear science to practical implementations of these systems. Machine learning methods are only effective in situations where the dynamics of the system are quite complex and not fully understood. Undoubtedly, we are at the point of an explosion of interest in the application of teaching methods in photonics. We hope our review will help the optical and laser communities become quickly familiar with this new field. We are happy to suggest interesting problems and to attract active students, physicists, mathematicians, and programmers, who are interested in this topic to join our teams.