Search

In the era of covid-19, the response in controlling covid-19 requires an adequate supply of medicines in hospitals and pharmacies. The supply of these drugs is felt to be very important because seeing the impact of the covid-19 virus is very dangerous for human health. Therefore, the existence of the supply of medicines in pharmacies during this Covid era really needs to be considered. Every transaction of the sale of the drug in pharmacies is always recorded. This sales transaction data can be processed to find certain patterns in selling drugs in a certain period of having so many drug sales transaction activities. If the sales transaction data is analyzed, a pattern can be found out that is very helpful in estimating drug stocks from the drug sales data. To be able to estimate drugs, the right method is needed in order to get accurate results. One of them is the Fp-Growth method. With Fp-Growth, data is entered for calculations based on drug sales data so as to produce conclusions of minimum support and minimum confidence values. Based on this study, 35 data on drug sales transactions were taken. The test results of this study obtained calculation results from simultaneous drug sales data that are often purchased by consumers with the highest accuracy are 17.14% and 50%, namely Paracetamol and CTM.

Most retail companies have implemented computer-based information systems for recording sales transaction data. In the implementation of information systems, the data collected in the database is processed limited to making reports such as sales reports and inventory reports. Database generated from computer-based information systems can be further processed to obtain more valuable information. One strategy for using sales transaction data is to analyze consumer spending patterns. Consumer spending patterns can be in the form of associations of items that are often purchased simultaneously. The association between goods can be determined using the frequent itemset search technique. The Fp-growth algorithm is an algorithm that can be used to determine frequent itemsets in a data set. This article describes the results of implementing the FP-Growth algorithm in the consumer shopping pattern analysis application. The resulting shopping pattern is in the form of goods that are often purchased simultaneously by consumers. From the results of the application of the fp-growth algorithm, it was found that the minimum value of support had an effect, namely the smaller the input value of support, the more pairs of items were obtained. The application of the FP-Growth algorithm in determining frequent itemsets in association data mining can find customer spending habits in buying goods simultaneously.

Nowadays the competition between companies is increasing. Companies need to predict their customers to find out the level of customer loyalty. One way is to analyze customer data by doing Customer Churn Prediction. In this study the method used is the FP-Growth Algorithm. The FP-Growth algorithm is an algorithm that uses the association rules technique to determine the data that appears most frequently. The data used in this study are secondary data and have 7,403 data from customers. The data has 21 variables. By using a minimum support of 1.2% and confidence at 80%, the associative rules generated are 60. The variable of the type of internet the customer has is strong enough to predict churn. It can be seen that of the 60 associative rules, there are 36 associative rules that have this variable. Testing associative rules on test data yields an accuracy of 71%.

Bakery shop is a shop that sells variants menu like bread, cakes, and drinks. The main problem with this store's sales is still not knowing which product items are best sellers and the shop still markets a lot of non-selling menus, causing the shop to lose money. So it takes the right strategy to increase the sales of bakery shop menus by making a menu package recommendations from the menus most frequently purchased by customers. The k-means algorithm performs grouping on menus to get menu packages. Furthermore, the fp-growth algorithm looks for linkages between frequently purchased menus to get menu package recommendations. The results of the research that the dominant items often purchased in cluster0 packages are hotdogs, pancakes, milk, garlic breadsticks with a confidence value of 92%, cluster1 packages are garlic breadsticks, hotdogs, chicken sand, pancakes with a confidence value of 92% and the last cluster2 packages are garlic breadstick, pastry, milk with a confidence value of 79%.