Artificial Intelligence and Applications in Research
Machine learning, neural networks and artificial intelligence are becoming almost inflationary buzzwords, nevertheless these techniques increasingly shape and influence science and society. Data-driven algorithms allow to cope with the tremendous amounts of data in modern experiments, ranging from particle [1] to astrophysics [2]. Especially for earth-observations there are also numerous examples where so-called deep learning networks [3] or a combination of different algorithms [4] can lead to new insights and open up new possibilities beyond established statistical methods.
In this talk I will be giving a brief introduction to modern data science methods and discuss some recent applications in science and research. I will highlight pros and cons of these algorithms and focus on examples from weather and climate research.
[1] Alexander Radovic et al., Nat. 560, 41-48 (2018)
[2] M. Garofalo et al., Proc. Int. Astr. U., 12(S325), 345-348 (2016)
[3] Markus Reichstein et al., Nature 566, 195–204 (2019)
[4] Thomas Weber et al., Nat. Com. 10, 4584 (2019)