Agricultural Data Mining Systems
Dashboards are analytic web applications that use data mining and machine learning techniques to explore a topic area (i.e. agriculture, health, infrastructure)
Our DMINE Data Portal uses Apache SOLR to organize data and repositories that are used for our modeling and dashboard analyses
We expose our models as JSON-accessible web service APIs, as well as via R packages
Our agricultural analysis approach explores crop insurance claim data (loss and claim frequency), in comparison to bioclimatic variables, over a broad set of commodities and damage causes. The USDA’s agricultural commodity loss archive (1980-present) is used as a key data source.
There is a known relationship between changing climate conditions and human health. Our work in this area uses known climatological variables, in concert with other associated feature variables, to construct a climate health predictive model.
Physical infrastructure is a huge expenditure in many aspects of society. Given these costs, impacts given climate change can have far reaching effects. Our work proposes to predict infrastructure resilience under changing climate regimes.