thieu1995/mealpy-timeseries

Hyper-parameter tuning of Time series forecasting models with Mealpy

Hyper-parameter tuning using Mealpy of:

1. MLP
2. SVR (SVM)

Notes: Run too slow

1. MLP: Can't complain, it takes time to train a single neural network

2. SVR: https://stackoverflow.com/questions/47460201/scikit-learn-svr-runs-very-long

   • How did I know it took too long with SVR, when there is no print out option in SVR model like MLP.
   • Set up print() in generate_trained_model() function of timseries_svr.py

Environments

Open terminal in windows:

    python -m venv pve_mealpy_test
    pve_mealpy_test\Scripts\activate.bat
    pip install -r requirements.txt
    

Cite Us

If you are using mealpy in your project, we would appreciate citations:

@software{nguyen_van_thieu_2022_6684223,
  author       = {Nguyen Van Thieu and Seyedali Mirjalili},
  title        = {{MEALPY: a Framework of The State-of-The-Art Meta-Heuristic Algorithms in Python}},
  month        = jun,
  year         = 2022,
  publisher    = {Zenodo},
  version      = {v2.4.2},
  doi          = {10.5281/zenodo.6684223},
  url          = {https://doi.org/10.5281/zenodo.6684223}
}