A methodology and life cycle model for data mining and knowledge discovery in precision agriculture

Presents a methodology for data mining and knowledge discovery in large, distributed and heterogeneous databases. In order to obtain potentially interesting patterns, relationships and rules from such large and heterogeneous data collections, it is essential that a methodology be developed to take advantage of the suite of existing methods and tools that are available for data mining and knowledge discovery in databases (KDD). One of the most important methodologies is an integration of diverse learning strategies that cooperatively performs a variety of discovery techniques that achieves high-quality knowledge. The KDLC (knowledge discovery life-cycle) model is an extended study of AqBC, which is a multi-strategy knowledge discovery approach that combines supervised inductive rule learning and unsupervised Bayesian classification via a constructive induction mechanism. A case study dealing with crop yields for a farm in the state of Idaho, USA is presented, and preliminary results are visualized by using the ArcView geographical information system (GIS). The significance of the multi-strategy knowledge discovery and visualization process in analyzing the classifications and learned rules has been empirically verified in KDLC.