Academia.eduAcademia.edu

Figure 3 – uploaded by bain khotimah

See full PDFdownloadDownload figure
Figure 3: Tree ending of item 68

Figure 3 Tree ending of item 68

Related Figures (10)

Each node in the FP-Tree contains three informations: item labeling, support count, and a connecting pointer that connects the nodes with the same item label between the paths. The picture below is the result of the formation of transactions 1 and 2. The role of FP-Growth Algorithm is to find the pattern of frequent itemset of FP-Tree that has been formed before. After FP-Tree is complete, FP- Growth algorithm looks for all possible subsets by generating conditional FP-Tree and searching for frequent itemset.
Each node in the FP-Tree contains three informations: item labeling, support count, and a connecting pointer that connects the nodes with the same item label between the paths. The picture below is the result of the formation of transactions 1 and 2. The role of FP-Growth Algorithm is to find the pattern of frequent itemset of FP-Tree that has been formed before. After FP-Tree is complete, FP- Growth algorithm looks for all possible subsets by generating conditional FP-Tree and searching for frequent itemset.
Table 8: Conditional Pattern Base and FP-Tree
Table 8: Conditional Pattern Base and FP-Tree
Figure 4: Tree after item 68 is omitted
Figure 4: Tree after item 68 is omitted
Table 10: Comparison of experimental time of Apriori algorithm and Fp Growth This suggests that the appearance of a combination of items more often does not necessarily lead to better recommendation results. Determination of variable data can determine the accuracy of FP-Growth made to determine minimum support and minimum confidence to find the interrelated frequent itemset as a strategy in the promotion of batik in SMEs. The highest value of precision is found on average confidence variations generated above 50%.
Table 10: Comparison of experimental time of Apriori algorithm and Fp Growth This suggests that the appearance of a combination of items more often does not necessarily lead to better recommendation results. Determination of variable data can determine the accuracy of FP-Growth made to determine minimum support and minimum confidence to find the interrelated frequent itemset as a strategy in the promotion of batik in SMEs. The highest value of precision is found on average confidence variations generated above 50%.
The calculation of lift ratio in Table 11, all the rules formed have a good strength because the value of the lift ratio rule is greater than 1. The calculation of lift ratio values is based on the assumption that consequent and antecedent are independent. The value of the lift ratio greater than 1 indicates that the attachment between two items, consequent and antecedent is strong, since the trust value or confidence rule that exceeds the minimum confidence is tested again on a percentage with the antecedent combination. Rule generated with the value of confidence close with min support 60% and min confidence 40%. They used the test for lift ratio (LR) with the result according to Table 11. Test results using Minimum Support = 60% and Minimum Confident 40% as follows:
The calculation of lift ratio in Table 11, all the rules formed have a good strength because the value of the lift ratio rule is greater than 1. The calculation of lift ratio values is based on the assumption that consequent and antecedent are independent. The value of the lift ratio greater than 1 indicates that the attachment between two items, consequent and antecedent is strong, since the trust value or confidence rule that exceeds the minimum confidence is tested again on a percentage with the antecedent combination. Rule generated with the value of confidence close with min support 60% and min confidence 40%. They used the test for lift ratio (LR) with the result according to Table 11. Test results using Minimum Support = 60% and Minimum Confident 40% as follows:
Table 11: Result on Calculation of LR of Batik SME Using Min Support 60% and Min Confidence 40%
Table 11: Result on Calculation of LR of Batik SME Using Min Support 60% and Min Confidence 40%
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved