IEEE Transactions on Evolutionary Computation
| Hi-index | 0.00 |
Grammatical Evolution (GE) is an Evolutionary Computing technique which can generate programs or codes in various languages based on the choice of a grammar. The evolutionary dynamics of GE is complicated and not well understood. The current body of knowledge on GE is largely based on empirical performance studies on some applications. There is little theoretical foundation or detailed analysis of evolutionary dynamics for GE in the literature. The limited knowledge on its mechanism is a limiting factor for applying GE to real world problems. An important real world application of data mining is the automated generation of knowledge from network intrusion data. Network intrusion detection systems are becoming a standard security feature in network infrastructures. Unfortunately current systems are not very good at detecting new types of intrusion without an associated high rate of false alarms. A goal of this research is to investigate and evaluate the real world application of data mining using GE, by assessing mechanisms for building effective and efficient intrusion detection systems based on GE. The methodology used involves fundamental theoretical analysis of GE, detailed analysis of its evolutionary dynamics and experimentation of GE concepts in mining datasets. The results include contributions to the body of scientific knowledge in Evolutionary Computing, GE and Data Mining.