Academia.eduAcademia.edu

Figure 2 – uploaded by bain khotimah

See full PDFdownloadDownload figure
Table-1. Comparison Of K-Means And GA+Kmeans Algorithm With Total Cost Fitness

Table 1 -1. Comparison Of K-Means And GA+Kmeans Algorithm With Total Cost Fitness

Related Figures (7)

This study was designed to SMEs parameter data grouping consisting of Total Labor, Investment Value, Value Production, Raw material type, value Raw Materials, Marketing. So, the GA+KMeans clustering methods performed to yield stable clusters using cluster centers point optimization with Genetic Algorithms. Then the results of the K-means clustering that has been in the GA optimization can form clusters that will be used to analyze the characteristics of the perpetrators of SMEs. Data will be processed to be in the cluster using the GA+KMeans and compared with the usual method of K-means. To measure the performance of each method used sum of square error (SSE). With this analysis is expected to help trade services selection policy in conducting training. SMEs need to be given special attention and assistance as appropriate. This study is expected to obtain a more optimal cluster centers that will assist in the formation of clusters which is more
This study was designed to SMEs parameter data grouping consisting of Total Labor, Investment Value, Value Production, Raw material type, value Raw Materials, Marketing. So, the GA+KMeans clustering methods performed to yield stable clusters using cluster centers point optimization with Genetic Algorithms. Then the results of the K-means clustering that has been in the GA optimization can form clusters that will be used to analyze the characteristics of the perpetrators of SMEs. Data will be processed to be in the cluster using the GA+KMeans and compared with the usual method of K-means. To measure the performance of each method used sum of square error (SSE). With this analysis is expected to help trade services selection policy in conducting training. SMEs need to be given special attention and assistance as appropriate. This study is expected to obtain a more optimal cluster centers that will assist in the formation of clusters which is more
Table-2. Eucledian Distance Fig.2. The Show is a graph comparing the number of clusters on SMEs that the results are not comparable. Tests conducted with 100 iterations between K-Means usual method with GA+KMeans. Results chart performance with the GA+KMeans more balanced than in the K-means ordinary. In these algorithms, every run generates a new solution so the values reported are averaged of 100 iteration..
Table-2. Eucledian Distance Fig.2. The Show is a graph comparing the number of clusters on SMEs that the results are not comparable. Tests conducted with 100 iterations between K-Means usual method with GA+KMeans. Results chart performance with the GA+KMeans more balanced than in the K-means ordinary. In these algorithms, every run generates a new solution so the values reported are averaged of 100 iteration..
Figure-2. Results SSE Of Chart Performance 100 Iteration: A. Group Cluster K-Means; B. Group Cluster GA+Kmeans Clustering method genetic algorithm and K-Means clustering can be used to categorize similar small industries in one or more regions, evidenced by mistake SSE < 0.1. The division of a dynamic group of SMEs conducted to examine the performance of the method, the results of the testing methods used to place the results of similar groups based on the accuracy of the most minimal. Groupped used as a reference to a policy in the development of SMEs to see the resemblance between the clusters. Results will be treated in the same cluster with sesame same cluster, so it will be easier for the government to provide assistance to increase SME.
Figure-2. Results SSE Of Chart Performance 100 Iteration: A. Group Cluster K-Means; B. Group Cluster GA+Kmeans Clustering method genetic algorithm and K-Means clustering can be used to categorize similar small industries in one or more regions, evidenced by mistake SSE < 0.1. The division of a dynamic group of SMEs conducted to examine the performance of the method, the results of the testing methods used to place the results of similar groups based on the accuracy of the most minimal. Groupped used as a reference to a policy in the development of SMEs to see the resemblance between the clusters. Results will be treated in the same cluster with sesame same cluster, so it will be easier for the government to provide assistance to increase SME.
Table-3. The Comparison Performance V Of Mean Intra- Cluster Distance (MICD) And Inter-Cluster Minimum Distance (ICMD)
Table-3. The Comparison Performance V Of Mean Intra- Cluster Distance (MICD) And Inter-Cluster Minimum Distance (ICMD)
Tabel-4. The Result Intra-Cluster Distance Within 50 Iterasi Table -5. Comparison SSE Of K-Means And GA+Kmeans
Tabel-4. The Result Intra-Cluster Distance Within 50 Iterasi Table -5. Comparison SSE Of K-Means And GA+Kmeans
The third test is done with 50 iterations shows the cluster results of both methods. Table 4. Shows the value of SSE for 50 iterations. Values are the smallest SSE In all clusters found on_ the GA+KMeans method. It can be concluded that the method has a good performance with test parameter: and p.mutasi GA with p.crossover higher then it car affect the outcome of the process of grouping
The third test is done with 50 iterations shows the cluster results of both methods. Table 4. Shows the value of SSE for 50 iterations. Values are the smallest SSE In all clusters found on_ the GA+KMeans method. It can be concluded that the method has a good performance with test parameter: and p.mutasi GA with p.crossover higher then it car affect the outcome of the process of grouping
REFRENCES: Figure 3. Graph above is a graph comparing the performance of SSE using five iterations between K-means usual method with GA-KMeans. Charts for all the smaller cluster GA+KMeans of the K-
REFRENCES: Figure 3. Graph above is a graph comparing the performance of SSE using five iterations between K-means usual method with GA-KMeans. Charts for all the smaller cluster GA+KMeans of the K-
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved