Every brain is different and so are their functions and dysfunctions. In recent years, large-scale recordings of neural activity have provided increasingly large datasets, and the scientific questions and models used to address them have grown in size and complexity. Our goal is to obtain an effective systems-level description of relevant neural dynamics in the context of cognitive functions and dysfunctions. Building on the foundations of dynamical systems and stochastic processes, we study the appropriate language for neural dynamics that can explain and generate specific predictions on neural data and behavior. To arrive at a model of neural computation tightly tied to the biology, we work closely with experimental and clinical collaborators. We develop statistical and machine learning methods for analyzing spatiotemporal neural and non-neural time series to infer neural dynamical systems. To facilitate the scientific inference process and clinical device development, we develop real-time machine learning and control methods for next-generation experiments.