AI prelims

Intelligence

The ability to acquire and apply knowledge benefit from past experiences act decisively to solve problems and adapt to new situations.

Artificial Intelligence (AI)

A branch of computer science concerned with designing intelligent systems that exhibit human-like intelligence represent knowledge symbolically and solve complex problems effectively and safely.

John McCarthy

The individual who coined the term Artificial Intelligence.

Linguistic Intelligence

Spoken and written language skills.

Logical-Mathematical Intelligence

Number skills and logical reasoning ability.

Musical Intelligence

Performance or composition skills in music.

Spatial Intelligence

The ability to evaluate and analyze the visual world.

Bodily-Kinesthetic Intelligence

Dance movement or athletic ability.

Interpersonal Intelligence

Skills in understanding and relating to others.

Intrapersonal Intelligence

Skills in understanding oneself.

Naturalist Intelligence

Skills in understanding the natural world.

Triarchic Theory of Intelligence

A theory proposed by Robert Sternberg consisting of practical creative and analytical intelligence.

Practical Intelligence (Contextual)

A person's ability to function effectively in a daily changing environment.

Creative Intelligence (Experiential)

A person's ability to adapt to new situations produce new ideas and handle novel to automated tasks.

Analytical Intelligence (Componential)

An individual's academic problem-solving ability based on standard IQ tests.

Aristotle

Formulated laws governing the rational mind and developed syllogisms for mechanical reasoning from initial premises.

Ramon Llull

Devised 'Ars Magna' (The Great Art) a reasoning system using rotating paper wheels.

Thomas Hobbes

Suggested reasoning as numerical computation (adding/subtracting) and proposed the idea of a thinking machine.

René Descartes

Discussed the distinction between mind and matter (dualism) and debated free will versus purely physical minds.

David Hume

Proposed the principle of induction in 'A Treatise of Human Nature' stating general rules come from repeated associations.

Rudolf Carnap and Carl Hempel

Analyzed knowledge acquisition from experience using confirmation theory and logical belief degrees.

Boolean Logic

Formal propositional logic developed by George Boole.

Gerolamo Cardano

First framed the idea of probability in terms of outcomes of betting events.

Ronald Fisher

Considered the first modern statistician who unified probability experimental design data analysis and computing.

Alan Turing

Characterized computable functions and established foundational principles of computability.

Intractable Problem

A problem where the time required to solve instances grows exponentially with the size of the instance.

Neuroscience

The study of the nervous system and the brain's information processing mechanism.

Paul Broca

Initiated the study of the brain's functional organization by investigating speech deficits (aphasia).

Camillo Golgi

Developed the staining technique allowing the visual observation of individual neurons.

Hans Berger

Invented the electroencephalograph (EEG) to measure intact brain activity.

Functional Magnetic Resonance Imaging (fMRI)

Neuroimaging technology providing detailed images of real-time cognitive brain activity.

Psychology

The science or study of the mind and behavior.

Hermann Von Helmholtz

Applied scientific methods to study vision and argued that perception involves unconscious logical inference.

Wilhelm Wundt

Opened the first experimental psychology laboratory to study thought processes through introspection.

Cognitive Psychology

A field viewing the brain as an information-processing device traceably linked to William James.

Kenneth Craik

Reestablished 'mental' terminologies (beliefs/goals) as legitimate scientific concepts in 'The Nature of Explanation'.

Donald Broadbent

Modeled psychological phenomena as information processing in 'Perception and Communication'.

Doug Engelbart

Pioneer of HCI who championed intelligence augmentation (computers aiding rather than replacing humans).

Computational Linguistics

A hybrid field intersecting computer science and linguistics to process natural human languages.

Noam Chomsky

Published 'Syntactic Structures' introducing formal structures addressing creativity in language.

Problem Solving

The early AI research area aiming to implement human-like problem-solving procedures in computers.

Game Playing

An AI implementation area using structured board games (e.g. chess checkers) due to simple state representation.

Natural Language Processing (NLP)

AI domain focused on creating systems that understand interpret and respond in human language.

Robotics

The science associated with designing fabricating and programming machines to perform specific physical tasks.

Speech Recognition

AI capabilities enabling systems to hear and comprehend spoken sentences accents and dialects.

Agent

Anything that perceives its environment through sensors and acts upon that environment using actuators.

Sensor

A device that detects physical inputs or conditions within an environment (e.g. camera GPS heat sensor).

Actuator

A mechanism that produces motion or action based on a detected stimulus (e.g. motors gears steering).

Percept

The specific content or input an agent's sensors perceive at any given moment.

Percept Sequence

The complete history of everything an agent has ever perceived.

Agent Function

An abstract mathematical function mapping any given percept sequence to an action.

Agent Program

The concrete physical implementation of an agent function running on an artificial system.

Performance Measures

Criteria used to evaluate how successful or desirable a sequence of environment states generated by an agent is.

Rational Agent

An agent that selects actions expected to maximize performance measures based on percept sequence and prior knowledge.

State Space

A set of possible states that an environment can be in often represented as a graph where vertices are states and directed edges are actions.

Initial State

The starting state that the agent begins in.

Goal State

The target state or set of states that an agent aims to reach.

Transition Model

A description of what each specific action does within a state space.

Action Cost Function

A function reflecting a problem-solving agent's performance measure by assigning a cost to each action.

Standardized Problem

A benchmark problem with a concise exact description intended to test and compare algorithm performance.

Real-world Problem

A problem with idiosyncratic formulations whose solutions are utilized directly by the general population.

Problem-solving Agent

An agent that considers a sequence of actions forming a path to a specific goal using atomic state representations.

Planning Agent

An agent that utilizes structured representations of different states to achieve goals.

Abstraction

The process of removing details from a representation while retaining validity and execution ease.

Ignorable Problem

A problem classification where solution steps can be ignored such as in theorem proving.

Recoverable Problem

A problem classification where solution steps can be undone such as in the 8-puzzle game.

Irrecoverable Problem

A problem classification where solution steps cannot be undone such as in checkers.

Heuristic Search

An experiential search process involving techniques to solve complex problems where no general algorithms exist.

State Space Representation

A problem representation method defining problems as positions at certain times to plan search algorithms.

Forward State Space Search

Also known as progression planning; starts at the initial state and moves forward to the goal state.

Backward State Space Search

Also known as regression planning; moves backwards from the goal state to the starting stage.

8-Puzzle Problem

A problem involving a 3x3 grid with eight numbered tiles and one empty space moved to reach a specified target layout.

Water Jug Problem

A classic problem requiring a specific volume of water using jugs of fixed capacities through filling pouring and emptying.

Problem Reduction

A technique that reduces a complex problem into simpler sub-problems often visualized using AND-OR graphs.

AND-OR Tree

A graphical representation of problem reduction where AND nodes require solving all sub-problems and OR nodes require solving any sub-problem.

Production System

A system mechanism providing rules (IF-THEN) to describe and execute search processes in automated planning and expert systems.

Production Rules

Rules consisting of a sensory precondition (IF) and an action (THEN) within a production system.

Working Memory

The central global database used by a production system to store current state information.

Control System

The component of a production system that selects applicable production rules and resolves execution conflicts.

Rule Applier

The core unit of a production system that executes the chosen production rules.

Specificity Conflict Resolution

A scheme that selects the rule with the most specific condition when multiple rules are triggered.

Recency Conflict Resolution

A scheme that favors the most recently added or matched rule in a production system.

Refraction Conflict Resolution

A scheme preventing a rule from firing again on unchanged working memory elements to avoid infinite loops.

Order Conflict Resolution

A conflict resolution scheme picking the first applicable rule based on its presentation sequence.

Completeness

An evaluation criterion measuring whether an algorithm guarantees finding a solution if one exists or reporting failure if not.

Cost Optimality

An evaluation criterion checking whether an algorithm finds the solution with the lowest path cost.

Time Complexity

A measurement of the time required by an algorithm to find a solution.

Space Complexity

A measurement of the memory required to run an algorithm.

Branching Factor (b)

The number of subsequent nodes or actions that need to be considered at any state in a search graph.

Depth (d)

The number of actions in an optimal solution path.

Maximum Path Length (m)

The maximum number of actions along any path in a state space.