ABSTRACT
The development of Knowledge Discovery in Databases (KDD) has been spurned by the exponential increase in digital data. One example is the large environmental data sets which will be generated by NASA’s Earth Observing System (EOS), including the Terra (formerly AM-1) and Landsat 7 satellites. One of the goals of these missions is to provide the capability to perform repeated global inventories of land-use and land-cover from space. Mapping land-cover from remote sensing has been an active area of research over the last twenty years. A variety of statistical parametric approaches as well as machinelearning approaches have been successfully applied to this problem.
