Filing system

Filing system

The goal of the classifier system is to optimize gain based on exposure to stimuli from a problem-specific environment. This is achieved by managing credit allocation for rules that prove useful and researching new rules and variations on existing rules using an evolutionary process.

Actors in the filing system include sensors, messages, effectors, comments, and classifiers. The detectors are used by the system to perceive the state of the environment. Messages are the information packets transmitted from the detectors to the system. The system performs information processing on the messages, and the messages can directly lead to actions in the environment.

Effectors control the actions of the system on and in the environment. In addition to the system actively perceiving via its detectors, it can also receive directed feedback from the environment (gain). Classifiers are condition-action rules that provide a filter for messages. If a message satisfies the conditional part of the classifier, the classifier's action is triggered. Rules act as message processors. A message is a fixed-length bit string.

A classifier is defined as a ternary string with an alphabet in {1, 0, #}, where the # represents whatever (corresponding to 1 or 0).

The system processing loop is as follows:

  1. Messages from the environment are placed in the message list.
  2. The conditions for each classifier are checked to see if they are met by at least one message in the message list.
  3. All satisfied classifiers participate in a contest, those who win display their action in the list of messages.
  4. All messages directed to effectors are executed (causing actions in the environment).
  5. All messages in the message list of the previous cycle are deleted (messages persist for only one cycle).

Binder systems are suited to problems with the following characteristics: perpetually new events with high noise, ongoing real-time demands for action, implicit or inaccurate goals set, and sparse gains or reinforcements that can only be achieved through long sequences of tasks.

The learning rate for the expected gain, error, and fitness of a classifier is usually in the range [0.1; 0.2]. The frequency of execution of thegenetic algorithm must be in the range [25; 50]. The discount factor used in multi-step programs is usually around 0.71. The minimum error that classifiers are considered to have equal precision is usually 10% of the maximum reward. The probability of crossing in the genetic algorithm is generally of the order of [0.5; 1.0]. The probability of mutating a single position in a workbook in the genetic algorithm is usually between [0.01; 0.05].

The experience threshold during classifier suppression is usually around 20. The experience threshold for a classifier during subsumption is usually around 20. Initial values for expected gain, error and adequacy of a classifier are usually small and close to zero. The probability of selecting a random action for exploration is generally close to 0.5. The minimum number of different actions that must be specified in a match set is usually the total number of possible actions in the environment for the entry.

Subsumption should be used on problem domains that contain well-defined rules for mapping inputs to outputs.

filing system