Clustering Macroeconomic Time Series

Abstract

The data mining technique of time series clustering is well established in many fields. However, as an unsupervised learning method, it requires making choices that are nontrivially influenced by the nature of the data involved. The aim of this paper is to verify usefulness of the time series clustering method for macroeconomics research, and to develop the most suitable methodology.By extensively testing various possibilities, we arrive at a choice of a dissimilarity measure (compression-based dissimilarity measure, or CDM) which is particularly suitable for clustering macroeconomic variables. We check that the results are stable in time and reflect large-scale phenomena such as crises. We also successfully apply our findings to analysis of national economies, specifically to identifying their structural relations.JEL: E00, C18, C63

Proceedings of Blockchain in Kyoto 2021 (BCK21)

“…However, in DTW, the data sequences are warped non-linearly in the time dimension to see a similar pattern allowing for lags and leads. DTW is a widely applied algorithm in non-economic fields, such as speech pattern recognition, though some studies have already used DTW in economics, for example, [2], [3], and [4]. Based on the DTW-based similarity measure, we find that the similarity in daily returns of cryptocurrencies with other assets has been present even since 2016, the beginning of our sample, and not much has changed throughout our sample period.…”

“…For example, in the case of BTC, we set these parameters as n = 1420 (i.e., q = 145), T = 19, f = 0.0526. This roughly means that we take out short-run cycles occurring for less than 20 business days, which is essentially one month in the calendar 2 .…”

Classical Machine Learning Methods in Economics Research: Macro and Micro Level Examples

“…Unsupervised Learning methods are widely used in various fields of economic research. For example, in the research of the relationship between financial market structure and the real economy [17], macroeconomic time series [18], the regional competitiveness [19]. The rule engine methods used in researches [20].…”

Biomass Clusterization from a Regional Perspective: The Case of Lithuania

“…Researchers base the reasons for the use of the K-mean differently. The K-Mean clustering algorithm is a basic algorithm based on the decomposition method used for many clustering tasks, especially for small-sized data sets [67]. Using the K-mean clustering method, data are grouped according to their proximity to each other according to Euclidean distance.…”