Fig. 5
From: Comparative analysis of deep learning architectures in solar power prediction
Training and validation loss curves (Mean Squared Error) for various deep learning architectures applied to solar power forecasting: (a) Transformer, (b) LSTM, (c) GRU, (d) CNN, (e) InformerLite, (f) SimpleRNN, (g) Temporal Convolutional Network (TCN), and (h) Autoencoder.
