Application of Genetic Algorithm K-Means Clustering of Villagers Characteristics For Smart Economy
Eneng Tita Tosida, Indra Permana Solihin, Fajar Delli Wihartiko, Farhan Naufal · 2022 International Conference on Informatics, Multimedia, Cyber and Information System (ICIMCIS) · 2022
One of the solutions carried out in helping rural economic growth is to implement a smart economy in smart village ecosystem. The purpose of this study is to apply the K-Means Genetic Algorithm grouping technique to analyze the readiness of villagers of Kabandungan Sub-regency Sukabumi Regency for the smart economy. In research, there is also a concept that is useful in the course of the research process, namely KDD. Tests were carried out on 350 respondents and 221 variables. Three methods are applied to the data namely GA, K-Means and GA K-Means. The optimal cluster value obtained at the elbow is 3 / K=3. The Davies Bouldin index (DBI) value is 14,511 for the GA method, next DBI value is 2,608 for the K-Means method, and the last value is 2,472 for the GA K-Means method. This makes the selection of the GA K-means method further analyzed, then the ANOV A approach is carried out to see the significance of the method results. The clustering results show that there are 23% of residents who are not ready for the smart economy, 57% of citizens who are sufficiently prepared for the smart economy, and 20% of residents who are very prepared for the smart economy. and the population of Cianaga Village who occupy the readiness is very well prepared & quite well prepared. The researchers' findings show that there are 3 levels of citizen participation, namely where cluster 1 is the crowdsourcing level, then cluster 2 distributed intelligence, and cluster 3 participatory science.