Skip to content

A comprehensive comparison of recent developed meta-heuristic algorithms for streamflow time series forecasting problem

License

Notifications You must be signed in to change notification settings

aiir-team/physics_MLP-monthly-streamflow-code

Repository files navigation

A comprehensive comparison of recent developed meta-heuristic algorithms for streamflow time series forecasting problem

Links:

  1. Code https://github.com/aiir-team/physics_MLP-monthly-streamflow-code
  2. Paper: https://doi.org/10.1016/j.asoc.2021.107282

Models

Traditional

  1. MLP (Streamflow forecasting using artificial neural network and support vector machine models)
  2. RNN (https://doi.org/10.1007/s11600-019-00330-1)

Evolutionary

  1. GA-MLP (Predicting monsoon floods in rivers embedding wavelet transform, genetic algorithm and neural network)
  2. DE-MLP (Differential evolution algorithms applied to neural network training suffer from stagnation)
  3. FPA-MLP (Not found yet, you can check again)

Swarm

  1. PSO-MLP (Hybrid particle swarm and neural network approach for streamflow forecasting)
  2. WOA-MLP (Annual Rainfall Forecasting Using Hybrid Artificial Intelligence Model: Integration of Multilayer Perceptron with Whale Optimization Algorithm)
  3. GWO-MLP (Improving artificial intelligence models accuracy for monthly streamflow forecasting using grey Wolf optimization (GWO) algorithm)
  4. SpaSA-MLP (No one - A novel swarm intelligence optimization approach: sparrow search algorithm)

Physics (Our proposed)

  1. WDO-MLP (Wind Driven Optimization - No one use it yet, but this algorithm is old - 2010)
  2. MVO-MLP (Multi-Verse Optimization - No one use it yet, 2016)
  3. EO-MLP (Equilibrium optimizer: A novel optimization algorithm - No one use it yet, 2019)
  4. NRO-MLP (Nuclear Reaction Optimization: A novel and powerful physics-based algorithm for global optimization - No one use it yet, 2019)
  5. HGSO-MLP (Henry gas solubility optimization: A novel physics-based algorithm, 2019)

How to run code

  1. Run MLP model by: mlp.py
  2. Run Rnn-based model (RNN, LSTM) by: script_traditional_rrn_based.py (multiprocessing)
  3. All hybrid-MLP model by: script_hybrid_mlp.py (multiprocessing also)
  4. Get the results table mean, std, cv by: get_table_results.py
  5. Results saved in folder: history/results/ based on daily or weekly datatype

Results in paper

  • Convergence errors of models in docs/results_of_paper/final_results_from_Thieu.xlsx
  • Stability figures in the paper in docs/results_of_paper/images/...

About

A comprehensive comparison of recent developed meta-heuristic algorithms for streamflow time series forecasting problem

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages