Fig 1.
Location of case study.
Table 1.
Basic information on the carbon emissions dataset.
Fig 2.
Flowchart of MDL model.
Table 2.
ADF test results of industrial carbon emission series in Xinjiang.
Fig 3.
Basic structure of LSTM.
Table 3.
Emission factors for different energy varieties.
Table 4.
Hyperparameter settings during model training.
Table 5.
Prediction results of different methods for industrial carbon emissions in Xinjiang.
Table 6.
Number of model training iterations.
Fig 4.
Comparison of performance of different models on RMSE.
Fig 5.
Comparison of performance of different models on MAE.
Fig 6.
Comparison of performance of different models on SMAPE.
Table 7.
Mean values of evaluation indicators of each model under different prediction ranges.
Fig 7.
Performance comparison between MAML_LSTM and MDL.
Table 8.
Percentage of performance increase or decrease in each metric for D_LSTM vs. DML.
Fig 8.
Comparison between LSTM and MAML_LSTM in terms of number of iterations.