Unsupervised Machine Learning Using Fp-Growth in Service and Maintenance Of Asset Management

Abstract

Preventive maintenance is one of effort in manufacturer industry to maintain an infrastructure that has an important thing in the industry. One of module that was provided from this system, like service and maintenance or Work Order (WO). This module has behavior data like brain jobs in human beings. Where is the data that record in memory used to learn to get a solution in the next experienced because the data WO be saved like data in the market basket analysis. The transaction data may be repeated in the next problem. So this data is interesting to be processed to get the best solution by involved it as machine learning like the recommended solution in the brain of human beings. This research will be focused on using fp-growth association rule as unsupervised machine learning to process data as a recommended solution for the technician. A different method like previous research using apriori algorithm. This research has a goal to prove the effect of minimum support with the result of decision support in fp-growth algorithm. The study shows the best condition of the result in this method is between 0.002 until 0.004 for minimum support, because the best precision ,recall, and accuration value more than 50% in that range of minimum support.