Method Overview
Random embeddings serve as a baseline control method to evaluate the effectiveness of learned representations. By generating random vectors for each light curve, we establish a lower bound for performance that helps demonstrate the value of sophisticated feature learning approaches.
