Activity Dependence

Activity dependence, the influence of prior activity on subsequent behavior in systems, is a central theme across diverse scientific fields, aiming to understand and model how past events shape current responses. Current research focuses on developing novel probabilistic models, such as discrete kernel point processes, to better control and quantify positive and negative dependencies, and on improving the explainability of complex models like deep neural networks by identifying regions of input data responsible for model competency. These advancements are crucial for enhancing the reliability and interpretability of machine learning models in various applications, from financial marketing to environmental sound analysis, and for gaining a deeper understanding of complex dynamical systems.