ai chapter 1
TRANSCRIPT
-
7/30/2019 AI Chapter 1
1/27
-
7/30/2019 AI Chapter 1
2/27
Knowledge : is the awareness and understanding of facts, truths or information gained inthe form of experience or learning through introspection.
Artificial Intelligence aims to improve machine behavior in tackling such complex tasks.
1.2 Branches of AI
Over the past five decades, AI research has mostly been focusing on solving specificproblems.
Numerous solutions have been devised and improved to do so efficiently and reliably. Thisexplains why the field of Artificial Intelligence is split into many branches, ranging fromPattern Recognition to Artificial Life , including Evolutionary Computation and Planning .
The idea is to develop an intelligent systems that has human level intelligence or better.
Natural Language Processing: to enable machines to successfully communicate in Englishlike languages, understanding the domain of the text.
Knowledge Representation : Facts about the world have to be represented in some way.Usually languages of mathematical logic are used.
Common sense knowledge and reasoning : This is the area in which AI is farthest from
human-level, in spite of the fact that it has been an active research area since the 1950s.While there has been considerable progress, e.g. in developing systems of non-monotonicreasoning and theories of action, yet more new ideas are needed.
Machine Learning : to adapt to new circumstances and to detect and extrapolatepatterns.
Computer vision : The world is composed of three-dimensional objects, but the inputs tothe human eye and computers' TV cameras are two dimensional. 2
-
7/30/2019 AI Chapter 1
3/27
1.3 Some Application areas of AISpeech
synthesisRobotics
ExpertSystems
Vision Systems
Naturallangauage
TheoremProving
ApplicationArea of AI
Expert Systems : AI programs that achieve expert-level competence in solving problemsin task areas by bringing to bear a body of knowledge about specific tasks are calledknowledge-based or expert systems .
Genetic algorithms Genetic Algorithms are adaptive heuristic search algorithmpremised on the evolutionary ideas of natural selection and genetic.
Neural Networks: A neural network is an interconnected group of neuron s.And others.
3
http://encyclozine.com/Neuronhttp://encyclozine.com/Neuron -
7/30/2019 AI Chapter 1
4/27
AI as a systemthat act rationally
AI as a systemthat act humanly
AI as a systemthat thinkrationally
Concerned withthought
processing andreason
AI as a systemthat thinkhumanly Concerned
withbehavior
4
1.4 What is AI(Artificial Intelligence)Different Scholars define AI differently
-
7/30/2019 AI Chapter 1
5/27
Thinking humanly: Cognitive modelingRequires scientific theories of internal activities of the brain.How to validate? Requires
1) Predicting and Testing behavior of human subjects(top-down).2) Direct identification from neurological data(bottom-up)
Study on mental processing in logic of human being is not yet fertile.AI as system that think humanlythe automation of activities that we associate with human thinking, activities such as decision making,
problem solving, learning . AI as system that act humanlythe act of creating machines that perform functions that require intelligence when performed by
people.
Thinking rationally
Right thinking is related to irrefutable reasoning process.Require structure that always gave correct conclusion given correct premises.Logic is the key to design and implement an agent that think rationally.Several Greek schools developed various forms of logic: notation and rules of derivation for thoughts.Direct line through mathematics and philosophy to modern AI.Problems:
1. Not all intelligent behavior is mediated by logical deliberation(unable totake informal knowledge for decision making process)
2. What is the purpose of thinking? What thoughts should I have?( there is abig difference between being able to solve a problem in principle anddoing so in practice)
5
-
7/30/2019 AI Chapter 1
6/27
AI as system that think rationally
A system is said to be rational if it does the right thing given what itknows.
The study of mental faculties through the use of computational models. The study of the computational that make it possible to perceive ,reason,and act.
AI as system that act rationally computational intelligence is the study of the design of intelligent agent. AI..is concerned with intelligent behavior in artifact.
Acting rationally : rational agentMeans acting so as to achieve ones goals, given ones belief. In this approach, AI is viewed as the study and construction of rational agent.Rational behavior. Doing the right thingThe right thing: that which is expected to maximize goal achievement. Giventhe available information.Doesnt necessary involves thinking: - e.g blinking reflex .-But thinking should be in the service of rational action.One way of acting rationally is to reason logically to the action. This indicates,making correct inference is part of being a rational agent.But rationality doesnt require correct inference because sometime withouthaving correct thing to do, agent must act rationally.
6
-
7/30/2019 AI Chapter 1
7/27
1.5 Foundations of AI The big question: How does the mind arise from the
brain?
Philosophy , logic, methods of reasoning
Mathematics formal representation and proof algorithms,computation, (un)decidability, (in)tractability, probability
Psychology adaptation, perceptionLinguistics knowledge representation, grammar
Neuroscience physical substrate for mental activity
Control theory simple optimal agent designs and so on.
Reading assignment
History of AI and different types of system
7
-
7/30/2019 AI Chapter 1
8/27
2. Intelligent AgentAn agent is anything that can be viewed as perceiving its environment through sensorsand acting upon that environment through actuators.
Assumption: Every agent can perceive its own actions (but notalways the effects)
Human agent:o Eyes, ears, and other organs for sensorso Hands, legs, mouth, and other body parts for actuators
Robotic agent:o Cameras and infrared range finders for sensors;o Various motors for actuator
It is also useful to think of intelligent systems as being agents, either:With their own goals or That act on behalf of someone (a user) An agent is an entity that exists in an environment and that acts on that environmentbased on its perceptions of the environmentAn intelligent agent acts to further its own interests (or those of a user).
An intelligent agent is a computer system capable of flexible autonomous action insome environment
By flexible , we mean: reactive pro-active social
8
-
7/30/2019 AI Chapter 1
9/27
ReactivityIf a programs environment is guaranteed to be fixed, the program neednever worry about its own success or failure program just executes blindly
Example of fixed environment: compiler
The real world is not like that: things change, information is incomplete.Many (most?) interesting environments are dynamic
Software is hard to build for dynamic domains: program must take intoaccount possibility of failure ask itself whether it is worth executing!
A reactive system is one that maintains an ongoing interaction with itsenvironment, and responds to changes that occur in it (in time for theresponse to be useful)
Proactiveness
Reacting to an environment is easy (e.g., stimulus response rules) But we generally want agents to do things for us Hence goal directed behavior Pro-activeness = generating and attempting to achieve goals; not driven
solely by events; taking the initiative Recognizing opportunities
9
-
7/30/2019 AI Chapter 1
10/27
10
HOW AGENTS SHOULD ACT A rational agent is one that does the right thing for the
perceived data from the environment. Even though doing right thing is ambiguous concept it is the
one that will cause the agent to be most successful. Thatleaves us with the problem of deciding how and when toevaluate the agents success.
-
7/30/2019 AI Chapter 1
11/27
Agents and environments
The interaction of agent and its environmentPercept: agents perceptual input at any given instant Percept sequence: complete history of everything the agent has ever perceived An agents choice of action at any given instant can depend on the entire
percept sequence observed to date An agents behavior is described by the agent function which maps from
percept histories to actions:[f: P* A]
We can imagine tabulating the agent function that describes any given agent(External characterization)
Internally, the agent function will be implemented by an agentprogram which runs on the physical architecture to produce f agent = architecture + program 11
-
7/30/2019 AI Chapter 1
12/27
Require more flexible interaction with the environment, theability to modify ones goals, knowledge that be appliedflexibly to different situations.
Degrees of Intelligence Building an intelligent system as capable as humans
remains an elusive goal. However, systems have been built which exhibit various
specialized degrees of intelligence. Formalisms and algorithmic ideas have been identified as
being useful in the construction of these intelligentsystems.
Together these formalisms and algorithms form the
foundation of our attempt to understand intelligence as acomputational process. In this course we will study some of these formalisms and
see how they can be used to achieve various degrees of intelligence.
12
-
7/30/2019 AI Chapter 1
13/27
Vacuum-cleaner world
13
One simple function is :if the current square is dirty then suck, otherwise move to the other square
-
7/30/2019 AI Chapter 1
14/27
Rational agents Rational agent is one that act rationally. It is action is measured to get
its performance.Performance measure: Subjective and Objective
Subjective measure :E. g. How happy is the agent at the end of action. Subjective measure is not better
Objective measure : needs standard to measure success. E.g., performance measure of a vacuum-cleaner agent could be
amount of dirt cleaned up, amount of time taken, amount of electricity
consumed, amount of noise generated, etc. As a general rule, it is better to design performance measures
according to what one actually wants in the environment. Rather thanaccording to how one thinks the agent should behave (amount of dirtcleaned vs a clean floor)
A more suitable measure would reward the agent for having a cleanfloor
Omniscience agent is different from rational agent. It is an agent thatknows the actual outcome of its action.
14
-
7/30/2019 AI Chapter 1
15/27
Environment types Fully observable vs. partially observable Deterministic vs. stochastic Episodic vs. sequential Static vs. dynamic Discrete vs. continuous Single agent vs. multiagent
Environment typesFully observable vs. partially observable: An environment is fully observable if an agent's sensorsgive it access to the complete state of the environment ateach point in time. Fully observable environments are convenient, because theagent need not maintain any internal state to keep track of the world
An environment might be partially observable because of noisy and inaccurate sensors or because parts of the stateare simply missing from the sensor data Examples: vacuum cleaner with local dirt sensor, taxidriver
15
-
7/30/2019 AI Chapter 1
16/27
Deterministic vs. stochastic: The environment is deterministic if the next state of the environment is
completely determined by the current state and the action executed bythe agent.
In principle, an agent need not worry about uncertainty in a fullyobservable, deterministic environment.
If the environment is partially observable then it could appear to bestochastic. E.g Vacuum world is deterministic while taxi driver is not
If the environment is deterministic except for the actions of other agents,then the environment is strategic
Episodic vs. sequential : In episodic environments, the agent's experience is divided into atomic
"episodes" (each episode consists of the agent perceiving and thenperforming a single action), and the choice of action in each episodedepends only on the episode itself. Examples: classification tasks
In sequential environments, the current decision could affect all futuredecisions. Examples: chess and taxi driver
16
-
7/30/2019 AI Chapter 1
17/27
Static vs. dynamic: The environment is unchanged while an agent is deliberating. Static environments are easy to deal with because the agent need not
keep looking at the world while it is deciding on the action or need it
worry about the passage of time. Dynamic environments continuously ask the agent what it wants to do. The environment is semi -dynamic if the environment itself does notchange with the passage of time but the agent's performance scoreDoes. Examples: taxi driving is dynamic, chess when played with a clock is
semi-dynamic, crossword puzzles are staticDiscrete vs. continuous: A limited number of distinct, clearly defined states, percepts andactions.
Examples: Chess has finite number of discrete states, and hasdiscrete set of percepts and actions. Taxi driving has continuousstates, and actions
17
-
7/30/2019 AI Chapter 1
18/27
Single agent vs. multiagent: An agent operating by itself in an environment is
single agent. Examples: Crossword is a singleagent while chess is two-agents
Question: Does an agent A have to treat an objectB as an agent or can it be treated as astochastically behaving object
Whether B's behaviour is best described by asmaximizing a performance measure whose valuedepends on agent's A behaviour
Examples: chess is a competitive multiagentenvironment while taxi driving is a partiallycooperative multiagent environment
18
-
7/30/2019 AI Chapter 1
19/27
SummaryEnvironment types
Chess with Chess without Taxi drivinga clock a clock
Fully observable Yes Yes NoDeterministic Strategic Strategic NoEpisodic No No No
Static Semi Yes NoDiscrete Yes Yes NoSingle agent No No No
The environment type largely determines the agent desig The real world is (of course) partially observable, stochastic, sequential,
dynamic, continuous, multi-agent
Reading assignment about Agent Application and PEAS
-
7/30/2019 AI Chapter 1
20/27
Structure of intelligent agentAgent functions and programs An agent is completely specified by the agent function mapping percept
sequences to actions One agent function (or a small equivalence class) is rational Aim: find a way to implement the rational agent function concisely ->
design an agent program. Agent = agent program + architecture Architecture: some sort of computing device with physical sensors and
actuators (PC, robotic car)should be appropriate: walk action requireslegs.
Agent functions and programsAgent program: Takes the current percept as input from the sensors
Return an action to the actuators While agent function takes the whole
percept history, agent program takes just the current percept as inputwhich the only available input from the environment The agent need toremember the whole percept sequence, if it needs it.
20
-
7/30/2019 AI Chapter 1
21/27
Table-lookup agentA trivial agent program: keeps track of the percept sequence
and then uses it to index into a table of actions to decide
what to do. The designers must construct the table that contains theappropriate action for every possible percept sequence.
Function TABLE-DRIVEN-AGENT(percept) returns an actionstatic : percepts, a sequence, initially empty table, a tableof actions , indexed by percept sequences, initially fullyspecified append percept to the end of percepts
action
-
7/30/2019 AI Chapter 1
22/27
-
7/30/2019 AI Chapter 1
23/27
Simple reflex agents
-
7/30/2019 AI Chapter 1
24/27
Model-based reflex agents
-
7/30/2019 AI Chapter 1
25/27
Goal-based agents
-
7/30/2019 AI Chapter 1
26/27
Utility-based agents
-
7/30/2019 AI Chapter 1
27/27
Learning agents