Please enable JavaScript.
Coggle requires JavaScript to display documents.
Unit 5: Artificial Intelligence - Coggle Diagram
Unit 5: Artificial Intelligence
The Beginnings of AI
Renowned Attempt at AI; Frankenstein, to create life
Modern History; John McCarthy the topic of the Dartmouth Conference (artificial intelligence)
HAL ( Heuristically programmed ALgorithmic computer)
capable of speech recognition, natural language understanding, lip reading, and thinking well enough to beat humans at chess
higher cognitive functions; kills its astronaut crew
central core; a room full of brightly lit computer modules mounted in arrays from which they can be inserted or removed
Focus of Intelligence in AI
ability to employ knowledge
ability to generate new ideas/ conceive new persprectives
ability to drawing conclusions
process of acquiring knowledge, skills, experience/values
the inner knowledge
Understanding AI
thought processes and reasoning
system thinking rationally
laws of thought-logic
interruptions, distractions, emotions
system thinking humanly
theory of the mind (human thought processes)
behavior
system acting rationally
an agent; perceives and acts, autonomous =, interacts with other agents plus the environment, reactive to the environment, pro-active(goal-directed)
system acting humanly
Turing Test( Alan Turing ); distinguishing intelligent entities from unintelligent ones; required of intelligence machine (human-like capabilities)-natural language processing, knowledge representation, automated reasoning, machine learning
Historical Background
The Jacquard Loom; ancestor of computer
Analytical Engine( Babbage; father of the modern day computer)-calculating machine, Mark 1 (the world's first program-controlled calculator
Turing's Proposal; Automated Computed Engine
aspects; erasable memory, convert binary form of calculations to decimal form of input and output, logical control, central arithmetic
advantages; speed, avoid human error, carry out complicated task
Turing Machine
Turing Test
John W Mauchly & J.Presper Eckert-Electronic Numeral Integrator and Calculator (ENIAC); Von Neumann Architecture.
Electronic Discrete Variable Automatic Computer (EDVAC)
Electronic Discrete Variable Automatic Computer (EDVAC)'Ballistic Research Laboratories. Operational until Dec 1962
The Machine Performs ( Modelling the Brain )
Early Programs
Human Logic
Logical steps
Logic Theorist Program
Newell, Shaw and Simon,1956; problem solving skills
proving thirty-eight of the first fifty-two theorems (Chapter 2 of Principia Mathematica) ; could reason
The General Problem Solver (GPS)
a set of operators
beginning at the initial state (beginning condition)
moves the process along step by step until problem reach solution
Expert Systems
encode human expertise in limited domains, a computer program designed to hold the accumulated knowledge of one or more domain experts
How Does it Work
knowledge base; facts, inference rules
inference engine; rule interpreter
user interface (workspace); problem description, problem status
Applications
PUFF; medical system for diagnosis of respiratory conditions
PROSPECTOR; used by geologists to identify sites for drilling/mining
MYCIN; medical system for diagnosing blood disorders
DESIGN ADVISOR; gives advice to designers of chips processor
DENDRAL; used to identify the structure of chemical compounds
LITHIAN; gives advice to archaeologists examining stone tools
Issues
The need
not always available
human experts are not 100% reliable and consistent
experts may not be good at explaining decisions
cost effective
Problem
limited domain
systems are not always up to date and don't learn
no common sense
need to setup and maintain system
Natural Language Processing
system that translate ordinary human natural language into language computers ,can understand and act on
challange
sense of context that humans have
human language filled with ambiguity and require context for understanding
attempt
ELIZA (1965); Joseph Weizenbaum
simulated a conversation between a patient& a psychotherapist
SCRIPT APPLIER MECHANISM (SAM); Roger Schank
program designed to understand stories that rely heavily on scripts
SEMANTIC INFORMATION PROCESSING (SHRDLU); Terry Winogard
program for understanding natural language (teletype-the blocks world)
ability to manipulate toy blocks
twelve most commonly used letters in the English language
carries out command given by human
Machine Translation
automatic translation
input (source language)
output (target language)
memory (words of source language and their equivalent in target language)
difficulty with translation- meaning
Machine Learning
Artificial Neural Network (ANN)
a system based on the operation of biological neural networks
simplest possible 'node' is a perceptron
two inputs
one output
applications
fields of identification and control system
game-playing AI & decision making
pattern recognition
face
object
voice
sequence recognition
health diagnosis
E-mail spam filtering
Robotics
essential features
sensing
light, hearing, touch pressure, chemical, taste
movement
wheels, legs, thrusters
energy
solar, electricity, battery
intelligence
programming
applications
animat; artificial life
project Cog