Experimental study of a hybrid intelligent system for flood risk management

This work carries out the experimental study of a neuro-fuzzy-genetic hybrid framework and demonstrates its potential, strengths and capabilities in flood risk management. A six layered neuro-fuzzy inference engine was formulated where the first, second and fifth layers consist of adaptive nodes while the third, fourth and sixth layers are fixed nodes. The perception of emergency risk management is very important in modern society; therefore this work demonstrates its practical application, data mining techniques and tools for emergency risk management. The implementation of knowledge extraction by rule discovery, rule evaluation and rule pruning is carried out. The work made used of MatLab programming language and Tanagra data mining software as front engines while Ms-Access database served as back engine. The evaluation of the proposed hybrid intelligent system using flood data obtained from Nigeria Emergency Management Agency (NEMA) is carried out. The research examined the application of hybrid intelligent for descriptive and predicative analytic framework in the domain of flood risks management with good result in flood data clustering, visualization, classification and predication.