User Tools

Site Tools


T-720-ATAI-2016 Main


Lecture Notes, F-3 19.01.2016


Minimal agent sensory data → decision → action
Perception Transducer that turns energy into information representation.
Decision Computation that uses perceptual data; chooses one alternative over (potentially) many for implementation.
Action Potential of the Agent to influence its task-environment, e.g. to move its body, grasp an object, utter some words, etc. Decisions turned into Actions produces Behavior.
Learning agent Uses memory to enhance decisions.

An abstraction of an agent: An agent has an input (i, selected from a task-environment), current state (S), goal (G, implicit or explicit), and output (o) in the form of atomic actions (selected from possible output), and a set of processes (P).

Complexity of Agents

Agent complexity Determined, at a minimum, by iXPXo, not just P, i, or o.
Taking time into account, as we should, an adaptive agent is of course more complex than the minimum (in the case of human-level intelligence, much more complex).
Agent action complexity potential Potential for P to control combinatorics of, or change, o, beyond initial i (at “birth”).
Agent input complexity potential Potential for P to structure i in post-processing, and to extend i.
Agent P complexity potential Potential for P to acquire and effectively and efficiently store and access past i (learning); potential for P to change P.
Agent intelligence potential Potential for P to coherently coordinate the above to improve the agent's ability to use its resources, or acquire more resources, to achieve top-level goals.

Reactive Agent Architecture

Architecture Largely fixed for the entire lifetime of the agent.
super simple Sensors connected directly to motors, e.g. Braitenberg Vehicles.
simple Deterministic connections between components with small memory, e.g. chess engines, Roomba vacuum cleaner.
Complex Grossly modular architecture (< 30 modules) with multiple relationships at more than one level of control detail (LoC), e.g. speech-controlled dialogue systems like Siri.
Super complex Large number of modules (> 30) at various sizes, each with multiple relationships to others, at more than one LoC, e.g. subsumption architecture.

Braitenberg Vehicle Examples

Braitenberg vehicle example control scheme: “love”. Steers towards (and crashes into) that which its sensors sense.
Braitenberg vehicle example control scheme: “hate”. Avoids that which it senses.
Braitenberg vehicle example control scheme: “curious”. The thinner wires are weighted-down signals, changing the behavior of “love” by avoiding crashing into things.

Subsumption Examples

Subsumption control architecture building block.
Example subsumption architecture for robot.
Subsumption architecture example, level 0.
Subsumption architecture example, level 1.
Subsumption architecture example, level 2.

Model-Acquiring Agents

Model A model of something is an information structure that behaves in some ways like the thing being modeled.
A good model of X …allows What-If questions to be asked about X, with answers correctly predicting what happens to X.
Model acquisition The ability to create models of (observed) phenomena.

Reinforcement Learning

Download Jordi's slides here, and see reading and study materials here.

2016©K. R. Thórisson


/var/www/ailab/WWW/wiki/data/pages/public/t-720-atai/atai-16/lecture_notes_f-3_19.01.2016.txt · Last modified: 2018/08/27 07:57 by thorisson