Abstract: A system and method for collective data mining from a distributed, vertically partitioned feature space as described. Collective data mining involves a unique approach for finding patterns from a network of databases, each with a distinct feature space. A distributed data mining system from heterogeneous sites is described. The architecture is ideal for accommodating different inductive learning algorithms for data analysis at different sites and includes a scalable approach using a gene expression-based evolutionary algorithm. This approach is used for distributed fault detection in an electrical power distribution network. Further implementations are also described.