no code implementations • 10 Jul 2022 • Han Yue, Steve Xia, Hongfu Liu
META consists of Positional Encoding, Transformer-based Autoencoder, and Multi-task Prediction to learn effective representations for both migration prediction and rating prediction.