Time Series Shapelets Classification Method for Predicting Global Temperature Anomalies

Accurate prediction of global temperature anomalies is crucial for understanding climate change and informing policy decisions. This paper presents a novel time series classification method based on shapelets for predicting global temperature anomalies. Shapelets are small, discriminative subsequences within time series data that capture important patterns. By leveraging shapelets, our method identifies and utilizes these patterns to improve classification accuracy. We propose a framework that extracts and selects relevant shapelets from historical temperature data, followed by a classification algorithm to predict future anomalies. Experimental results demonstrate that our shapelet-based approach significantly outperforms traditional time series classification methods in terms of predictive accuracy and robustness. This work provides a valuable tool for climate scientists and policymakers to better anticipate and respond to global temperature changes.