Implementation of K-Means and Agglomerative Hierarchical Methods to House Clusterization

Abstract

People in general will think that building a house with a larger building area will cost more than building a house with a smaller building area. This view is not always correct because one of the factors that affect the size of the cost depends on the size of the building and the ratio of the mixture between cement, sand and lime. The smaller the cement mixture for buildings, the smaller the costs will be. Based on this case, the researchers grouped the data (clustering) using Euclidean distance to measure the distance between points. Grouping 200 data based on 2 features, namely the amount of cement and the amount of costs that do not have a label. The results showed that clustering with the K-Means method was able to group 200 data into 3 groups with the results of group one as many as 50 data, group 2 as many as 50 data, group 3 as much as 100 data with a computation time of 0.444 seconds and silhouette 0.82 while the results of clustering research using the Agglomerative Hierarchical method with a single linkage shows 100 data in group 1,50 data in group 2, 50 data in group 3 with a computation time of 3.22 seconds and silhouette 0.51