Customized dynamic load balancing for a network of workstations
Journal of Parallel and Distributed Computing
Mining Very Large Databases with Parallel Processing
Mining Very Large Databases with Parallel Processing
Knowledge Discovery in Databases
Knowledge Discovery in Databases
Efficient Attribute-Oriented Generalization for Knowledge Discovery from Large Databases
IEEE Transactions on Knowledge and Data Engineering
Database Mining: A Performance Perspective
IEEE Transactions on Knowledge and Data Engineering
Extending Attribute-Oriented Induction as a Key-Preserving Data Mining Method
PKDD '99 Proceedings of the Third European Conference on Principles of Data Mining and Knowledge Discovery
Discovery of multiple-level rules from large databases
Discovery of multiple-level rules from large databases
| Hi-index | 0.00 |
Attribute-Oriented Induction (AOI) reduces the search space of large data to produce a minimal rule set. Classical AOI techniques only consider attributes that can be generalised but eliminates keys to relations. The Key-Preserving AOI (AOI-KP) preserves keys of the input relation and relate them to the rules for subsequent data queries. Previously, the sequential nature of AOI-KP affected performance on a single processor machine. More significantly, time was spent doing I/O to files linked to each generated rule. AOI-KP is O (np) and storage requirement O (n), where n and p represent the number of input and generalised tuples respectively. We present two enhanced AOI-KP algorithms, concAOI-KP (concurrent AOI-KP) and onLineConcAOI-KP of orders O (np) and O (n) respectively. The two algorithms have storage requirement O (p) and O (q), q = p*r, 0r≤ l respectively. A prototype support tool exists and initial results indicate substantially increased utilisation of a single processor.