ARTIFICIAL INTELLIGENCE UNIT 4

123456789101112131415161718
Across
  1. 5. Instantiations of some or all of the random variables describing the domain
  2. 6. Network that has connections only in one direction
  3. 11. Technique that combats overfitiing
  4. 14. Inspired by William of Ockham
  5. 15. This process of explicitly penalizing complex hypotheses
  6. 17. The proportion of mistakes an hypothesis makes
  7. 18. Gives the highest information gain
Down
  1. 1. A network with all the inputs connected directly to the outputs
  2. 2. Representation in which the states are defined by set of features
  3. 3. Algorithm that resembles decision-list learning
  4. 4. A computer system modeled on the human brain and nervous system
  5. 7. The general task of learning a probability model
  6. 8. Each leaf has linear function of some subset
  7. 9. Gives a single output value after taking in a vector of attribute values
  8. 10. Takes a learning algorithm as an argument
  9. 12. Also called learning curves
  10. 13. Most common unsupervised learning task
  11. 16. Neural Networks are composed of