In supervised learning, we have features and class labels . Write a program that produces form

In unsupervised learning, we only have values, but no explicit target labels. We can

  • Outlier detection: is this a normal ?
  • Similarity/Clustering: which examples look like this
  • Which occur together
  • Latent-factors: what ‘parts’ are the made from
  • Data visualization: what does the high-dimension feature space look like?
  • Ranking: what are the most important ?