Lung Cancer is a disease in which there are cells that grow in the lungs by a collection of carcionogens uncontrollably. Lung Cancer can be treated with surgery, chemotherapy and radiotherapy. Early treatment that needs to be done to reduce the mortality rate in patients with lung cancer after performing thoracic surgery, by collecting data from each patients regarding this information causes a new problem, including the data obtained including high-dimensional data and has many attributes so that it can produce less accurate information. So it is necessary to calculate data mining clustering. In general, the methods for performing clustering are grouped into four parts, namely partitioning, hierarchial, grid-based and model-based. This study used the k-medoids algorithm because it is able to handle data sensitive to outliers and has high accuracy and efficiency in processing large numbers objects. The results of the k-medoids calculation were evaluated using the euclidean distance Davies Bouldin Index which resulted in a DBI value of 0.93543 indicating that the k-medoids algoritm achieves good grouping because the final result of the calculation is less than 0. From the results of the evaluation using DBI it shows that the k-medoids algorithm has an average accumulation average at the time of execution is quite fast and the cluster quality is good.

No abstract