/TS_RNN_Experiments

📉 Experiments with RNN prediction efficiency

Primary LanguageJupyter Notebook

TS_RNN_Experiments

Experiments setup to evaluate the performance of the RNN forecasting with different data processing method, forecasting strategies and cross-series methods

Dir tree

├── experiments  
│        ├── Compare_crossseries_info.ipynb - Experiment on the influence of cross-series methods of RNN forecasting on the quality of the RNN forecast  
│        ├── Compare_data_transformers.ipynb - Experiment on the influence of data processing methods on the quality of the RNN forecast  
│        └── Compare_prediction_methods.ipynb - Experiment on the influence of forecasting strategies on the quality of the RNN forecast  
├── reports  
│        ├── data_tr_experiment_all - outputs of data transformers experiment  
│        ├── factors_experiment_all - outputs of cross-series methods experiment  
│        ├── factors_experiment_all_notdeseason - outputs of cross-series methods experiment without deseason ts processing  
│        ├── strategy_experiment_all - outputs of forecasting strategies experiment  
│        └── strategy_experiment_all_notdeseason - outputs of forecasting strategies experiment without deseason ts processing  
├── Master_thesis.pdf - (rus) text of my master thesis
├── requirements.txt  
├── README.md  

Experiments

Compare_data_transformers Open In Colab

Compare_prediction_methods Open In Colab

Compare_crossseries_info Open In Colab