<![CDATA[The Consumer Price Index (IHK) is a value that calculates changes in the weighted average price of goods and services consumed by households and serves as the basis for the BPS-Statistics to calculate inflation. Weighting data from the BPS-Statistics cost of living survey (SBH) is one of the components required to explain and demonstrate the dynamics of the IHK. SBH is held in several cities due to the limited resources to conduct this survey. As a solution, the sister city approach is adopted by BPS-Statistics to estimate the consumer price index for cities that are not part of the cost-of-living survey domain. The sister city approach uses weighting data from a city that held SBH with similar consumption patterns and is located geographically close to each other. Although the appointment of a sister city went through several procedures, there was no existing method to measure how similar a city is to another city based on the sister city definition. In this paper, we will use machine learning to analyze the similarity of cities in Nusa Tenggara Timur based on their consumption patterns, and as a result, the decision to appoint a sister city will be more accurate. Machine learning is a field of artificial intelligence (AI) and computer science that uses data and algorithms to mimic how people learn and progressively increase its accuracy. Machine learning methods will support the sister city approach with scientific reasoning to produce more accurate inflation. The result of the clustering based on the elbow method for K-means shows that Kupang city has a unique characteristic which means there is no similarity with the other cities in Nusa Tenggara Timur. However, other cities grouped into two cluster where the two inflation cities (Maumere and Waingapu) are not in the same cluster.]]>
