Update dimensionality_reduction.py

machine_learning/dimensionality_reduction.py

@@ -183,14 +183,17 @@
  sb = covariance_between_classes(features, labels, classes)
  sw = covariance_within_classes(features, labels, classes)
 
+ # Add regularization to Sw to avoid singular matrix
+ sw_reg = sw + 1e-6 * np.eye(sw.shape[0])
+
  # Solve the generalized eigenvalue problem: Sb v = λ Sw v
- eigenvalues, eigenvectors = eigh(sb, sw)
+ eigenvalues, eigenvectors = eigh(sb, sw_reg)
 
  # Sort eigenvectors by eigenvalues (descending)
  idx = np.argsort(eigenvalues)[::-1]
  eigenvectors = eigenvectors[:, idx]
 
- # Take top "dimensions"
+ # Take top "dimensions" eigenvectors
  filtered_eigenvectors = eigenvectors[:, :dimensions]
 
  projected_data = np.dot(filtered_eigenvectors.T, features)
@@ -200,7 +203,6 @@
  logging.error("Dataset empty")
  raise AssertionError
 
-
 def locally_linear_embedding(
  features: np.ndarray, dimensions: int, n_neighbors: int = 12, reg: float = 1e-3
 ) -> np.ndarray: