Scientific Program
School Science Program
The school program will provide foundational and advanced knowledge of machine learning tools for research in modern physics through lectures on selected topics, including:
Statistics and uncertainty. Statistical methods (sampling methods, Markov Chains, MCMC, and Bayes approaches) are the foundation of modern ML algorithms. This school will provide a solid understanding of statistics and uncertainty, which is essential for ML applications.
Data Access. Data is the essence of ML and modern physics. This school will explore the best data acquisition and mining practices, data cleaning, and data preparation, drawing on academic and industry examples.
ML elements and foundations. The core part of the school will span from foundational to advanced notions on machine learning, including hot topics in the fast-evolving field of AI, offering a wide overview of the current landscape in the field and the instruments to understand and effectively use cutting-edge and future algorithms. Examples of applications of research in modern physics will be provided.
ML applications to biomedical data. ML methods have been applied to solve complex problems in many different fields, ranging from industry to healthcare and genomics. The school will provide insights into cutting-edge ML applications to high-throughput biological data, in particular in the context of single cell experiments. The focus will be on the methodological challenges posed by these data sets, emphasizing the need for interpretability to connect predictive power with scientific understanding.
Ample time will be devoted to hands-on tutorials (daily, Monday through Wednesday). Participants will also engage in a two-day hackathon (Thursday and Friday) to apply, in groups, the knowledge acquired during lectures to topical problems in industry and modern physics. To this end, while registering at the school, participants are encouraged to submit an open problem for the Hackathon sessions using this form: Call for Hackathon projects.
Participants at an early stage of their careers are welcome and encouraged to apply, and no significant background in AI is required. However, general fluency in Python coding is expected. For a productive experience, participants are encouraged to check if they possess the required background knowledge by consulting this page: Pre-exercises. A refresher tutorial is also available on this page: Scientific Computing with Python