Registration deadline for abstracts: Sunday February 23, 2025 [please submit the abstract in the "motivation section" upon registration].

Notification of accepted participants and talks: Friday March 7, 2025.

For this workshop, we propose to bring together researchers working on different aspects of stochastic processes, from theoretical developments to practical applications. This includes theory-oriented researchers demonstrating results in ideal configurations, whose findings can then be adapted to concrete cases studied in fields such as epidemiology, ecology and economics. It also includes application-oriented researchers who will motivate and discuss these practical cases. The workshop will address direct modeling of stochastic processes in concrete cases, and approaches to stochastic processes from an inference perspective. The aim is for these often separate communities to interact and enrich their respective research, by sharing insights from different perspectives and situations.

We plan to address specifically the following topics:

• Theory of Stochastic Processes

• Stochastic Inference

• Application of Stochastic Processes

Stochastic processes are ubiquitous in Nature and play a crucial role in the description of both physical phenomena and biological systems. They are equally important in modeling social interactions [1], where numerous agents are subject to individual changes that can be described as noise and interactions with their surroundings. These processes have vast practical implications. For example, they allow us to describe complex environments that we cannot fully control and whose description is not only a rich field for theoretical research, but also crucial for the proper handling of practical applications. A paradigmatic example is climate change [13], where understanding stochastic processes is essential for adapting our lifestyles and designing effective mitigation strategies. An important branch of stochastic processes, namely stochastic inference, has evolved considerably in recent years in the context of modeling and focuses on inferring the parameters that rule the stochastic dynamics in observed trajectories, whether the available data are fully or only partially observed [6].

In the last deceade, interest in stochastic processes and their applications has considerably grown in the scientific community. Theoretically, this has led to remarkable progress in the study of out-of-equilibrium dynamics using stochastic thermodynamics [22,23], the description of the extreme events statistics using record models [15], the physical description of avalanches in a wide variety of systems, or the characterization of fluctuations in random processes using large deviation theory [24,19].  On the other hand, stochastic approaches have become standard in numerous purely applied areas. Several examples illustrate this trend. In machine learning, neural networks are trained using the stochastic gradient descent algorithm, in which the gradients are averaged over a subset of training data, the minibatch. This process can be modeled by a Langevin process with a finite temperature [9,17], which provides valuable insights into the properties of the optimization problem and the identification of optimal algorithms. In epidemiology, the propagation of epidemics involves complex interactions between virus transmission and agents, which can be modeled by complex interaction graphs [20]. This approach allows researchers to study the reverse process of stochastic spread to trace back to the initial patient zero or zeros [1]. The financial world also relies heavily on stochastic modeling. For example, call options can be represented by random walks in simple cases so that it is possible to study how events and the agent's decisions influence stock prices [7].

The applications of stochastic processes are rapidly increasing with the increasing availability of data in the age of big data. Consequently, it is essential to integrate theoretical descriptions of stochastic processes with practical modeling efforts to address real-world problems in science and society. At this point, stochastic inference is of central importance as it facilitates the quantitative description of key observables through realistic modeling of complex phenomena, thus improving our ability to both understand and predict the behavior of complex systems.