Please enable JavaScript.

Coggle requires JavaScript to display documents.

Origin of RL, Optimal Control Describing the problem of designing a …

- - - - Integration
        
        Q Learning: Integration of MDPs & Learning Process by Watkins, 1989
- - - - Law of Effect: summarised the idea by Edward Thorndike, 1911
      - "Reinforcement" first appeared in the 1927
        English translation of Pavlov’s monograph on conditioned reflexes
      - Reinforcement is the strengthening of a pattern of behavior as a result of an animal receiving a stimulus—a reinforcer—in an appropriate temporal relationship with another stimulus or with a response. Some psychologists extended its meaning to include the process of weakening in addition to strengthening, as well applying when the omission or termination of an event changes behavior. Reinforcement produces changes in behavior that persist after the reinforcer is withdrawn, so that a stimulus that attracts an animal’s attention or that energizes its behavior without producing lasting changes is not considered to be a reinforcer.
        Quote from R.Sutton's Intro to RL
        
        Arthur Samuel (1959) was the first to propose and implement
        a learning method that included temporal-difference ideas
        
        Minsky (1954) Secondary reinforcers: a stimulus that has been paired with a primary reinforcer("reinforcement") such as food or pain and, as a result, has come to take on similar reinforcing properties
        
        1972, Klopf developed the idea of “generalized reinforcement,” whereby every component (nominally, every neuron) views all of its inputs in reinforcement terms
        
        He and Barto refined these ideas and developed a psychological model of classical conditioning based on temporal-difference learning (Sutton and Barto, 1981a; Barto and Sutton, 1982)
        
        Sutton (1978a, 1978b, 1978c) developed Klopf’s ideas further, particularly the links to animal learning
        theories
        
        Temporal Difference
        
        1 more item...
    - - Alan Turing's Pleasure-Pain System in 1948
      - Since Alan Turing's, researchers devised a lot of applications
        
        From this explorive interests to RL, Rosenblatt/Widrow/Hoff came up with Neural Networks and so on.. so RL -> supervised learning is the correct direction by R.Sutton intro to RL