Time Series Clustering Using Technical Indicators

Abstract

In this paper, we propose a method for clustering time-series data using technical indicators. Although a method for clustering companies by industry has been proposed for time-series stock price data, the clustering was performed on time-series data as independent features at each time, and the time-series nature of the data was not taken into account. The method proposed in this paper takes into account the time-series nature of the data, calculates technical indicators from the time-series data, and performs clustering using these indicators as features. By using technical indicators as features, clustering can be expected to take into account the time-series nature of the data. The proposed method is applied to actual stock price data and ECG data to confirm its effectiveness.