Wave-RVFL

Wave-RVFL: A Randomized Neural Network Based on Wave Loss Function

Please cite the following paper if you are using this code.

Reference: M. Sajid, A. Quadir, and M. Tanveer, "Wave-RVFL: A Randomized Neural Network Based on Wave Loss Function." Published in the 31st International Conference on Neural Information Processing (ICONIP) 2024.

Arxiv Link: https://arxiv.org/abs/2408.02824

Paper: https://doi.org/10.1007/978-981-96-6579-2_17

BibTex

---

@inproceedings{Sajid2025WaveRVFL,
  author    = {Sajid, M. and Quadir, A. and Tanveer, M.},
  title     = {Wave-RVFL: A Randomized Neural Network Based on Wave Loss Function},
  booktitle = {Neural Information Processing},
  series    = {Lecture Notes in Computer Science},
  volume    = {15287},
  pages     = {242–257},
  year      = {2025},
  publisher = {Springer},
  address   = {Singapore},
  doi       = {10.1007/978-981-96-6579-2_17}
}

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

The experimental hardware configuration includes a personal computer with an Intel(R) Xeon(R) Gold 6226R CPU at 2.90 GHz and 128 GB of RAM. The system runs Windows 11 and uses MATLAB 2023a to run all experiments.

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

We have provided a demo code for the "Wave-RVFL" model using the "adult" dataset.

For the demo purpose, the following hyperparameters were used for the experiment. To get the optimal results, please tune the hyperparameters. For the detailed experimental setup, please refer to the paper's supplementary material.

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Demo Hyperparameter setting

option.C=1; %Regularization parameter

option.N=810; %Number of hidden nodes

option.activation = 1; %Sigmoid Activation function

option.a=-1;

option.b=0.6;

option.alpha=0.0001;

option.m=2^8;

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Note: For detailed parameter settings, please refer to the "Wave-RVFL: A Randomised Neural Network Based on Wave Loss Function" paper.

Description of files: Wave_RVFL_Main.m: This is the main file to run selected models on datasets.

Wave_RVFL_Model.m: This is a model function used for training and testing.

sigmoid.m: Sigmoid activation function.

relu.m: ReLU activation function.

Selu.m: SeLU activation function.

Evaluate.m: Function to evaluate the accuracy.

adult.mat: adult dataset used to execute the code.

Remarks:

1. The codes have been cleaned for better readability and documented, then re-run and checked the codes only in a few datasets, so if you find any bugs/issues, please write to M. Sajid (phd2101241003@iiti.ac.in).

2. For the detailed experimental setup, please follow our paper.

Some parts of the code have been taken from:

1. M. Sajid, A. K. Malik, M. Tanveer and P. N. Suganthan, "Neuro-Fuzzy Random Vector Functional Link Neural Network for Classification and Regression Problems," in IEEE Transactions on Fuzzy Systems, vol. 32, no. 5, pp. 2738-2749, May 2024, doi: 10.1109/TFUZZ.2024.3359652.

2. Mushir Akhtar, M. Tanveer, Mohd. Arshad, "Advancing Supervised Learning with the Wave Loss Function: A Robust and Smooth Approach," in Pattern Recognition, Vol. 155, 110637, 2024, doi: 10.1016/j.patcog.2024.110637.