Implementation Machine Learning of K-Means Clustering Method and Linear Regression for Detecting the Risk of Tuberculosis Spread in Bangka Regency

The number of tuberculosis (TB) sufferers in Bangka Regency tends to increase and is becoming a very serious public health problem. In 2023, there will be around 331,581 residents of Bangka Regency, with around 1,489 of them suffering from TB. The high number of TB cases in Bangka Regency requires comprehensive monitoring. Surveillance can be carried out by grouping TB cases by region. One clustering method is K-Means which groups data based on similar criteria. Researchers also used the Linear Regression method to predict the effect of population density on increasing TB prevalence. This research was conducted to cluster the distribution pattern of TB by region and identify the influence of population density variables on the increase in TB. Based on the results of the K-Means method analysis by looking at the Performance Davies Bouldin-Index (Dbi) test, researchers obtained a pattern of TB distribution which was categorized into: Areas with High Prevalence (Sungailiat), Medium Prevalence (Belinyu and West Mendo) and Low Prevalence (Bakam, Merawang, Pemali, Pudding Besar, and Riau Silip). Meanwhile, the prediction results show that there is a positive influence of 81% of the population density variable on increasing the number of TB spreads and 19% is influenced by other variables. The results of this research can be used as a reference for the Bangka District Health Service to assist decision-making in preventing an increase in TB cases.