AI: Learning in AI 2
3 likes2,574 views
Bayesian learning views hypotheses as intermediaries between data and predictions. Belief networks can represent learning problems with known or unknown structures and fully or partially observable variables. Belief networks use localized representations, whereas neural networks use distributed representations. Reinforcement learning uses rewards to learn successful agent functions, such as Q-learning which learns action-value functions. Active learning agents consider actions, outcomes, and how actions affect rewards received. Genetic algorithms evolve individuals to successful solutions measured by fitness functions. Explanation-based learning speeds up programs by reusing results of prior computations.
AI: Learning in AI 2
- 1. Learning In AI 2
- 2. Bayesian learningBayesian learning views the problem of constructing hypotheses from data as a sub problem of the more fundamental problem of making predictions. The idea is to use hypotheses as intermediaries between data and predictions.
- 3. Belief network learning problems Known structure, fully observableUnknown structure, fully observableKnown structure, hidden variablesUnknown structure, hidden variables
- 4. A comparison of belief networks and neural networksThe principal difference is that belief networks are localized representations, whereas neural networks are distributed representationsAnother representational difference is that belief network variables have two dimensions of "activation"—the range of values for the proposition, and the probability assigned to each of those values.
- 5. What is Reinforcement learning?The task of reinforcement learning is to use rewards to learn a successful agent function.
- 6. What is Q-Learning?The agent learns an action-value function giving the expected utility of taking a given action in a given state. This is called Q-learning.
- 7. Passive learning in an unknown environmentThe prioritized-sweeping heuristic prefers to make adjustments to states whose likely successors have just undergone a large adjustment in their own utility estimates.
- 8. Active learning in an unknown environmentAn active agent must consider what actions to take, what their outcomes may be, and how they will affect the rewards received.
- 9. Design for an active Adaptive dynamic programming(ADP) agentThe agent learns an environment model M by observing the results of its actions, and uses the model to calculate the utility function U using a dynamic programming algorithm.
- 10. Exploration of Action An action has two kinds of outcome:It gains rewards on the current sequence.It affects the percepts received, and hence the ability of the agent to learn—and receive rewards in future sequences.
- 11. Genetic AlgorithmIt starts with a set of one or more individuals and applies selection and reproduction operators to "evolve" an individual that is successful, as measured by a fitness function. The genetic algorithm finds a fit individual using simulated evolution.
- 12. Knowledge from learningIf we use Descriptions to denote the conjunction of all the example descriptions, and Classifications to denote the conjunction of all the example classifications.The Hypothesis then must satisfy the following property: Hypothesis A Descriptions |= ClassificationsWe call this kind of relationship an entailment constraint.A cumulative learning process uses, and adds to, its stock of background knowledge over time.
- 13. Prior knowledge in learningIs useful in following waysBecause any hypothesis generated must be consistent with the prior knowledge as well as with the new observations, the effective hypothesis space size is reduced to include only those theories that are consistent with what is already known.For any given set of observations, the size of the hypothesis required to construct an explanation for the observations can be much reduced, because the prior knowledge will be available to help out the new rules in explaining the observations. The smaller the hypothesis, the easier it is to find.
- 14. Explanation based learningThe technique of memorization has long been used in computer science to speed up programs by saving the results of computation.
- 15. Improving efficiency in LearningA common approach to ensuring that derived rules are efficient is to insist on the operationality of each sub goal in the rule. EBL makes the knowledge base more efficient for the kind of problems that it is reasonable to expect.
- 16. Visit more self help tutorialsPick a tutorial of your choice and browse through it at your own pace.The tutorials section is free, self-guiding and will not involve any additional support.Visit us at www.dataminingtools.net