IDSV Chap. 11 - Artificial Intelligence

Heuristics

Mental shortcuts or "rules of thumb" that often lead to a solution (but not always). rules of thumb for making descisions

Current robots are successful on specific tasks:

Mobility (swim, fly, hop, etc) driving cars behave like pet dogs guide weapons to their targets

image analysis

understanding what the image characteristics represent in the real world

Contextual analysis

The context is brought into understanding process: "The bat fell to the ground"

intelligent agent

autonomous goal-directed entity which observes using SENSORS and acts upon an environment using ACTUATORS

How do artificial neuron networks learn?

From examples by adjusting their weight

Turing test

Human interrigator communicates with test subject by type writer, can the human interrogator distinguish wether the test subject is human or a machine?

Learning strategies

Learning by IMITATION SUPERVISED learning learning by REINFORCEMENT

Artificial Neuron

each weight is multiplied by a wrighting factor if sum of weighted inputs exeed theshhold then output is 1 else output is 0

Image processing steps

edge enhancement region finding

Semantic analysis

identifies the information content in a sentence: "Mary gave John a birthday card"/ "John got a birthday card from Mary" "Do you know what time it is?"

Syntactic analysis

indentifies the grammatical role of each word (parsing)

Requirements for good heuristics

it should constitute a reasonable estimate of proximity to a goal it should be easy to compute

Strong AI

machines can be programmed to possess intelligence and conciousness

Weak AI

machines can be programmes to exhibit intelligent behavior

Production systems

model of performance based on if-then {conditions-action} commands

learning by reinforcement

the agent is given a general rule to judge for itself when it has succeeded or failed

learning by imitation

the computer records the steps performed by a person

Robotics

the study of physical agents that behave intelligently

image processing definition

to change an image in order to emphasize the most important parts (identifying characteristics of an image)

Production system steps

1. collection of states -start/initial state -goal state 2. collection of productions (rules/moves) -from one state to another -might have preconditions 3. control system - decides what production to apply next

supervised learning

a person identifies the correct response for a number of examples and the agent generalizes from those examples

Semantic net

((pic))

search tree example

((pic))

state graph

((pic))

Language processing analyses:

Syntactic analysis Semantic analysis Contextual analysis