Applications

• I typically have a slide in each lecture that discusses applications to games.
• However, learning is ubiquitous, so it can be applied anywhere.
• Here are some particularly important ways of breaking up application domains:
  • Categorisation (supervised). See the University of California at Irvine's Categorisation Benchmark.
  • Classification (unsupervised)
  • Pattern Recognition
  • Prediction
• It strikes me that learning is not a technique that is used often in games, and that it is a way to move forward.