Use the Encoder-Decoder framework to build an LSTM network to complete the Seq2Seq prediction task.
The models folder contains trained models for testing purposes. Use trainer.py to train your own model, and use launch.py for prediction and testing. The main parameters of the network are configured in config.py under the configs folder. After you are familiar with the purpose of the parameters, you can change them to suit your own data.
tensorflow 1.13.1, numpy 1.16.0, pandas 0.24.2.
Data Sources：https://geodata.pku.edu.cn/index.php?c=content&a=show&id=728%E8%BF%99%E9%87%8C%E5%8F%AF%E4%BB%A5%E4%B8%8B%E8%BD%BD
Results of sea surface temperature prediction in some areas.
Heat map of global sea surface temperature in 2018.
