Understanding how Arabidopsis thaliana responds to cold stress at the metabolomic level is essential for uncovering plant resilience mechanisms. In this study, we applied Self-Organizing Maps (SOMs) for metabolomic prediction and pattern recognition. The dataset includes metabolite concentration values and realistic growth rates for 241 A. thaliana ecotypes, with each ecotype analyzed for 37 primary metabolites. These metabolites, particularly sugars, show significant concentration shifts in response to stress, making them ideal for detecting concept drift and understanding its impact on plant growth under cold stress conditions. The study utilized two distinct datasets: one from plants grown under standard growth conditions at 16 ℃, and the other from plants exposed to cold stress at 6 ℃. By applying SOMs to these data, we aimed to uncover patterns and predictive insights into the metabolomic changes induced by cold stress, providing new perspectives on the adaptive mechanisms of A. thaliana.