When to Use (Linear SVM - Non-Linear SVM - Logistic Regression)

  • 𝑛 - number of training examples
  • 𝑘 - number of features

if 𝑘 is large (relative to 𝑛) (e.g. 𝑘≥𝑛, 𝑘=10,000, 𝑛=10-1000)

  • use logistic regression or SVM without a kernel (linear kernel)

if 𝑘 is small, 𝑛 is intermediate (e.g. 𝑘=1-1000, 𝑛=10-10,000)

  • use Non-Linear SVM with Gaussian kernel

if 𝑘 is small, 𝑛 is large (e.g. 𝑘=1-1000, 𝑛=50,000+)

  • create/add more feature, then use logistic regression or SVM without a kernel
  • bc SVM is slow when number of training examples are large

neural network likely to work well for most of these settings, but may be slower to train