Analysing Annual Consumer Price Indexes Using the Clustering Algorithms

Abstract

Consumer price index is an important economic indicator for any country in the world. It makes a measurement of the change on the prices of a group of products in specific time intervals. Its importance is huge because it gives information on the measurement of inflation, indexed payments, pensions and not only. Different states have different annual consumer prices indexes and this has created differences among them. This article aims to examine and present clusters for annual consumer price indexes for some European countries using the Data Mining techniques. In this analysis are taken into account annual consumer price indexes of twenty-two countries for twenty-two years including Albania. The creation and analysis of clusters is carried out through algorithms of k-Means clustering, partitioning around medoids and hierarchical agglomerative clustering. Their results show the countries that have greater similarity of indexes in time and how similar are the results of each algorithm. The database processing and the presentation of the results is performed in R software.