Comparison of K-Means Clustering Algorithm with Fuzzy C-Means In Measuring Satisfaction Level Of Television Da'wah Surau TV
Abstract
Da'wah Television Surau TV is a broadcasting media that presents broadcasts around Islam. This media will quickly develop as it presents broadcasting material in meeting the spiritual needs of its viewers. To Increased media development is highly dependent on the satisfaction of the audience in all aspects of broadcast supporting. It is therefore, to measure the level of audience satisfaction as an effort to generate continuous broadcast quality improvement.This research is performing of algorithm clustering comparation with K-Means Clustering modeling and Fuzzy C-Means modeling to classify and mapping the most appropriate dataset so that it can assist analysing or measuring the level of audience satisfaction toward the da'wah television Surau TV. Comparison of clustering algorithm performance with K-Means Clustering modeling and Fuzzy C-Means modeling is based on processing speed and trace value of each RMSE parameter of clustering algorithm. The RMSE result of clustering research using algorithm with K-Means Clustering is 2.09879 and by using algorithm with Fuzzy C-Means model is 2.07911. Fuzzy C-Means modeling speed is faster in conducting the clustering process compared with K-Means Clustering modeling. It can be concluded that clustering with Fuzzy C-Means modeling is able to produce more accurate cluster compared to clustering with K-Means Clustering modeling accuracy Keywords: Clustering; K-Means; Fuzzy C-Means; Satisfaction rate survey; RMSE