SODIUM-ION BATTERY CYCLE LIFE PREDICTION USING MACHINE LEARNING

In this study, we explore the classification and prediction capabilities of three models-Genetic Programming (GP), Logistic Regression (LR), and the Kolmogorov-Arnold Network (KAN)-on the task of sodium-ion battery life prediction. By leveraging a dataset composed of multiple battery characteristics, we aim to determine the remaining power of sodium-ion batteries using these machine learning models. The KAN model, being a novel approach, demonstrates superior performance across various metrics, including accuracy, precision, recall, and F1 score, when compared to the other two models. This highlights the potential of KAN as a robust model for complex classification tasks in the field of battery life prediction.