<![CDATA[Tsunami is one of the most deadly disaster causing damage and loss of life and wealth. It happens in a sudden and unpredictable. Lack of awareness often leads to a great damage and worsening the impact of tsunami itself. This research implements genetic algorithm optimization into K-Means method for classify tsunami data. By optimazing the initial cluster center it will used as an input on K-Means method. The method result more optimal preference than the conventional K-Means method since the central point is optimized by genetic algorithm. It was proved on this research where fitness value resulted from Silhouette Coefficient to observe how suitable data with cluster. Chromosome representation used here is real code to initialize centroid value. Extended intermediate crossover applied for crossover method. For mutation method, random mutation is run here. Also for selection method it uses elitism selection. Based on testing result, the most optimum parameter accomplished are 50 population, 70 generation, and Cr =0.9 and Mr =0.1 combination with fitness value around 0.995934]]>
Copyrights © 2018
