Using Spatio-Temporal Analytics to Understand Seasonal Processes Seminar

Event details

In this presentation I will review a suite of spatio-temporal analytical approaches and their use to mine, model and map seasonal patterns and processes. More precisely, I will present and discuss both unsupervised and supervised learning approaches such as frequent pattern mining, co-clustering, and various machine learning methods. These approaches were recently used to analyze weather data, volunteered observations (plant phenology and tick bites) and very high spatial resolution satellite images. Results show recurrent wind patterns, years and locations with similar temperature and/or phenology, inconsistent phenological observations, environmental conditions associated to tick bites, spring onset at continental scales, and crop type maps of smallholder farmlands.