Last active
April 11, 2023 13:22
-
-
Save catorch/8ae4bd49b984b28c8837b4a92f39e51e to your computer and use it in GitHub Desktop.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| import numpy as np | |
| import torch | |
| import torch.nn as nn | |
| import torch.optim as optim | |
| from sklearn.model_selection import train_test_split | |
| from sklearn.metrics import accuracy_score, f1_score, classification_report | |
| class ELM(nn.Module): | |
| def __init__(self, input_dim, hidden_dim, output_dim): | |
| super(ELM, self).__init__() | |
| self.input_dim = input_dim | |
| self.hidden_dim = hidden_dim | |
| self.output_dim = output_dim | |
| # Initialize random weights for the input-to-hidden layer | |
| self.hidden_weights = nn.Parameter(torch.randn(input_dim, hidden_dim) * np.sqrt(2 / input_dim), requires_grad=False) | |
| # Initialize the hidden-to-output layer weights, which will be learned | |
| self.output_weights = nn.Parameter(torch.zeros(hidden_dim, output_dim), requires_grad=True) | |
| def forward(self, x): | |
| # Calculate the hidden layer output using the ReLU activation function | |
| h = torch.relu(x @ self.hidden_weights) | |
| # Calculate the output layer (no activation function) | |
| out = h @ self.output_weights | |
| return out | |
| def fit(self, x, y, alpha=1e-5): | |
| # Calculate the hidden layer output | |
| h = torch.relu(x @ self.hidden_weights) | |
| # Calculate the Moore-Penrose pseudo-inverse | |
| h_pseudo_inverse = torch.pinverse(h.T @ h + alpha * torch.eye(self.hidden_dim)) @ h.T | |
| # Calculate the optimal output weights | |
| self.output_weights.data = h_pseudo_inverse @ y | |
| def predict(self, x): | |
| with torch.no_grad(): | |
| out = self.forward(x) | |
| return torch.argmax(out, dim=1) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment