On the Applications of K-means Algorithm for Pattern Recognition in Large Data. An Informal Approach: Pattern Recognition using K-means Clustering a Comparative Study of COVID-19 Trade Data and Employment Index Data

This report presents a comparative study on the use of the K-means clustering algorithm for pattern recognition in two large, real-world datasets: COVID-19 trade impact data and employment index data. The objective is to examine how data characteristics such as volatility, stability, and the presence of outliers’ influence clustering quality and decision-making usefulness. Using recent literature, the study demonstrates that while K-means effectively structures both datasets into meaningful clusters, employment index data yields more stable and interpretable patterns suitable for long-term planning. In contrast, COVID-19 trade data, due to its highly dynamic nature, is more appropriate for short-term exploratory analysis. These findings align with recent advances in clustering research that emphasize the importance of data distribution and centroid stability (Awad & Hamad, 2022; Selmi et al., 2024) [2, 5].