Classification

Every classifier must be initialized with a specific set of parameters. Two distinct methods are deployed for the training and the testing phases. Whenever possible, the real valued prediction is stored in the realpred variable.

Binary Classification

Compute Model

compute(x, y)

x - training data [2D numpy array float]

• x.shape[0] number of samples
• x.shape[1] number of features

y - training classes (1 or -1) [1D numpy array integer]

• y.shape[0] number of samples

Test Model

predict(p)

p - test data [1D or 2D numpy array float]

• 1D: one sample
  • p.shape[0] number of features
• 2D: more than one sample
  • p.shape[0] number of samples
  • p.shape[1] number of features

Multiclass Classification

Compute Model

compute(x, y)

x - training data [2D float numpy array]

• x.shape[0] number of samples
• x.shape[1] number of features

y - training classes (1, ..., #classes) [1D integer numpy array]

• y.shape[0] number of samples

Test Model

predict(p)

p - test data [1D or 2D float numpy array]

• 1D: one sample
  • p.shape[0] number of features
• 2D: more than one sample
  • p.shape[0] number of samples
  • p.shape[1] number of features

Classifiers

• Support Vector Machines (SVMs)
• K Nearest Neighbor (KNN)
• Fisher Discriminant Analysis (FDA)
• Spectral Regression Discriminant Analysis (SRDA)
• Penalized Discriminant Analysis (PDA)
• Diagonal Linear Discriminant Analysis (DLDA)