The Implementation of Data Mining for Association Patterns Determination Using Temporal Association Methods in Medicine Data

Zahrotun, Lisna and Soyusiawati, Dewi and rahma sara, pattihua The Implementation of Data Mining for Association Patterns Determination Using Temporal Association Methods in Medicine Data. [Artikel Dosen]

Text The Implementation of Data Mining for Association Patterns.pdf Download (537kB)

Abstract

Abstract— Clinic is one of the businesses that perform health services for people in the surrounding environment. The clinic also provides medicines that will be given to patients who conduct health checks. The problem that occurs in these clinics is that the medicine data recap is only using excel data, the purchase of medicine stocks that are conducted only based on medicine that out of stock. Based on an interview with one of the nurse at a clinic on Yogyakarta site, occassionally, there are a case that a surge of patient that running out medicine supplies, while on the other hand there are lots of medicine accumulation occurred because these medicines was not needed by the patient. This is because the clinic has not been able to predict the medicine that are often issued by the clinic. Therefore, this research aims to build a data mining program with the Temporal Association Rules method for determining the relationship between medicines which is accompanied by the date of release of the medicine.The method used in this research is Temporal Association Rules with the Apriori Algorithm to find association rules that meet the support and confidence limits, and in the testing process lift ratio is used.The results of this research are applications that able to provide information on patterns of medicine data associations and the date of medicine’s release. The test results with 8186 amount of data and support value 50% and confident value 70% with lift values above 0, the patterns of association rules obtained is 6. these clinics is that the medicine data recap is only using excel data, the purchase of medicine stocks that are conducted only based on medicine that out of stock. Based on an interview with one of the nurse at a clinic on Yogyakarta site, there are 183 types of medicines with maximum of 630 times medicine expenditure in a month. Another problem that has not been resolved is medicines that are issued simultaneously for a certain time cannot be known. For example, during the rainy season in October- December, the expenditure of cough medicine (OBH) and paracetamol relatively increased sharply by 80% compared to normal days. As for the supply of medicines in stock globally. This condition shows some medicine items that are not needed in the excessive rainy season, while for other items of medicine are needed when the rainy season appears to be less in number. This is the impact of the problems that are owned by clinics that have not been able to find information on medicines issued simultaneously at a certain time. From the description above, in this research a pattern of associations was searched using the Temporal Association Rules method on medicine data. This research is expected to be able to find out the types of medicines that are often issued simultaneously with the time aspects in the results of association rules. Keywords— Data Mining, Apriori Algorithm, Temporal Association Rules, Lift Rasio, Medicine data

Item Type:	Artikel Dosen
Subjects:	T Technology > T Technology (General)
Divisi / Prodi:	Faculty of Industrial Technology (Fakultas Teknologi Industri) > S1-Informatics Engineering (S1-Teknik Informatika)
Depositing User:	Mrs. Lisna Zahrotun
Date Deposited:	07 Oct 2019 02:10
Last Modified:	07 Oct 2019 02:10
URI:	http://eprints.uad.ac.id/id/eprint/15181

Actions (login required)

View Item