Evolving Gaussian Processes and Kernel Observers for Learning and Control in Spatiotemporally Varying Domains: With Applications in Agriculture, Weather Monitoring, and Fluid Dynamics

Monitoring and modeling large-scale stochastic phenomena with both spatial and temporal (spatiotemporal) evolution by using a network of distributed sensors is a critical problem in many control applications (see 'Summary'). Consider, for example, a team of robots that has the task of destroying herbicide-resistant weeds on a farm (see Figure 1 and 'Key Control Problems in Agriculture'). This team must predict weed growth across the whole farm to make intelligent, coordinated decisions [1].