Parallel sessions

Recent years have seen a rise of data-driven research in several branches of astronomy and astrophysics as both hardware and software improvements have significantly facilitated the implementation of new statistical data analysis techniques. Multiple out-of-the-box toolkits such as tensorflow and scikit-learn are making machine learning more accessible than ever to non-specialists. However, the sheer volume of literature and overwhelming number of packages available makes it challenging for the self-taught researcher to find the right direction to approach these methods and to efficiently apply them to scientifically relevant questions in their field. The objective of this workshop is to provide beginners in machine learning with hands-on experience in implementing and using statistical-learning methods as well as background knowledge necessary to identify the additional tools they would require to solve specific astrophysical problems. In this two-hour tutorial, participants will implement a full machine-learning method from scratch and apply it to an astrophysical problem, while being given the opportunity to get a grasp of the potential and limitations of these statistical methods.