整理了CEEMDAN的多种组合模型的时间序列预测python代码合集,包含CEEMDAN,CEEMDAN-LSTM,CEEMDAN-CNN-BiGRU,CEEMDAN-CNN-BiLSTM,CEEMDAN-CNN-LSTM,CEEMDAN-CNN-BiLSTM-Attention多种分解组合预测模型。该算法表现优异,适合文章发表。
测试集指标 RMSE MAE MAPE R2
预测结果指标1:| 35.379559329900104 | 27.147625122070313 | 6.069660094287265% | 83.392523628316%
预测结果指标2: | 53.06147880467345 | 38.32430775960287 | 8.733205025723935% | 62.64424627122014%
预测结果指标3: | 45.646985727379516 | 33.98502655029297 | 8.153488815922138% | 72.3545913766921%
预测结果指标4: | 67.58365124383627 | 49.2855219523112 | 10.94421082218187% | 39.39867680264887%
预测结果指标5: | 32.66135202235171 | 26.006954650878907 | 6.042495746946329% | 85.846393352868%