Chapter 1-6 Flashcards by Jerome Domanico

This means doing the right thing

Rationality

How well did you know this?

Not at all

Perfectly

This type of AI talks about acting like human. Machine Learning, Computer Vision, NLP, Knowledge Presentation, Robotics

Acting Human

How well did you know this?

Not at all

Perfectly

This type of AI talks about how humans think. Psycholoogy, Neuroscience, Introspect

Thinking Human

How well did you know this?

Not at all

Perfectly

This type of AI talks about logic

Thinking Rational

How well did you know this?

Not at all

Perfectly

This type of AI talks about agent based AI

Acting Rational

How well did you know this?

Not at all

Perfectly

Give the four characteristics of AI

Human
Rational
Thinking
Act

How well did you know this?

Not at all

Perfectly

This pertains to an agent (entity) that do the right thing. It lis like a function

Rational Agent

How well did you know this?

Not at all

Perfectly

Give some risks of AI

Unemployment
Unmanned Warfare
Biased Decision Making
Cybersecurity
Security

How well did you know this?

Not at all

Perfectly

Give some benefits of AI

Time Efficiency
Easier surveilance

How well did you know this?

Not at all

Perfectly

These are entities with less rationality that makes them rational when worked with each other

Swarm

How well did you know this?

Not at all

Perfectly

An entity that has a percept which yields to an actuator

Agent

How well did you know this?

Not at all

Perfectly

This is also known as the problem. Everything outside the agent including other agents, physical objects, and digital inputs

Environment

How well did you know this?

Not at all

Perfectly

Devices that receives information about the environment

Sensors

How well did you know this?

Not at all

Perfectly

Devices that allow the agent to reflect to the environment, such as moving, speaking, or controlling digital interfaces

Actuators

How well did you know this?

Not at all

Perfectly

To define an AI program, you must define its four properties which are:

Performance
Environment
Actuators
Sensors

How well did you know this?

Not at all

Perfectly

Property of Task Environments: completeness of the information of the environment

Fully Observed vs Partially Observed

How well did you know this?

Not at all

Perfectly

Property of Task Environments: one or more agent

Single vs Multi agent

How well did you know this?

Not at all

Perfectly

In a multi agent setting, there are 2 properties of agent

Cooperative
Competitive

How well did you know this?

Not at all

Perfectly

Property of Task Environments: the next state of the environment is completely determined by the current state and action executed by the agents

Deterministic vs Nondeterministic

How well did you know this?

Not at all

Perfectly

Nondeterministic with Probability

Stochastic

How well did you know this?

Not at all

Perfectly

Property of Task Environments: current decision can affect future decisions

Sequential

How well did you know this?

Not at all

Perfectly

Property of Task Environments: doesn’t affect future decisions

Episodic

How well did you know this?

Not at all

Perfectly

Property of Task Environments: the environment doesn’t change

Static

How well did you know this?

Not at all

Perfectly

Property of Task Environments: the environment changes

Dynamic

How well did you know this?

Not at all

Perfectly

Property of Task Environments: there is exact position/state for entities in the environment

Discrete

Property of Task Environments: no exact position/state for entities

Continuous

Property of Task Environments: if the agent knows the environment, then it knows all the possible choices/outcomes already

Known vs unknown

This is a combination of architecture and program

Agent

Property of agent which consists of if-else

Simple Reflex Agent

Property of agent which creates a model (decision tree, rules, formulas)

Model Based Agent

Property of agent which focuses on attaining a goal

Goal Based Agent

Property of agent which is based on highest utility

Utility Based Agent

A machine learning type agent that uses model to train

Learning agent

This agent solves a problem by searching. Example is vacuum cleaner or self driving car that finds the shortest path to a destination

Goal Based Agent

How we think and act

Intelligence

Understanding but also building intelligent entities. Machines that can compute how to act effectively and safely in a wide variety of novel situation

Artificial Intelligence

This test asks a question: Can a machine think?

Turing Test by Alan Turing

Communicate successfully in human language

Natural Language processing

Stores what it knows or hears

Knowledge representation

Answers questions and to draw new conclusions

Automated reasoning

Adapt to new circumstances and detect patterns

Machine Learning

To pass the Turing Test, a robot must have these two:

Speech recognition, Computer Vision

Trying to catch our own thoughts as they go by

Introspection

Observing a person in action

Psychological Experiments

Observing a brain in action

Brain imaging

Combines models from AI and Psychology to construct precise and testable theories of human mind

Cognitive Science

Agent comes from the Latin word ____, which means to do

Agere

There are ___ foundations of AI

8 or Eight

Foundation of AI - study of reasoning

Philosophy

Foundation of AI - study of thinking

Psychology

Foundation of AI - logic

Mathematics

Foundation of AI - robotics, hardware

Computer Engineering

Foundation of AI - study of brain

Neuroscience

Foundation of AI - uncertainty

Economics

Foundation of AI - NLP

Linguistics

Foundation of AI - the way how you control things

Control Theory and Cybernetics

Set of possible states in an environment

State Space

This state is where the agent starts in

Initial State

True or False: An agent has only one goal state

False

True or False: Before an agent can start searching, a well-defined problem must be formulated

True

State the five parts of a problem

Initial State Actions Transition Model Goal States Action Cost Function

Search algorithms are judged based on _____, _______, _______, and _______.

Completeness Cost-optimality Time complexity Space Complexity

This search method only has access to the problem definition. Algorithms on this search will create a search tree to find the solution

Uninformed Search

This search has access to a heuristic function. They have access to additional information such as pattern databases with solution costs

Informed Search

______ selects nodes for expansion using an evaluation function. This is under _____ search

Best-First Search, Uninformed

______ expands the shallowest nodes first; it is complete, optimal for unit action costs, but has exponential space complexity. This is under _____ search

Breadth-first Search, Uninformed

______ expands the node with lowest path cost, g(n), and is optimal for general action costs. This is under _____ search

Uniform-cost search, Uninformed

______ expands the deepest unexpanded node first. This is under _____ search

Depth-first search, Uninformed

_____ is like depth-first search but with specific depth limits until a goal is found. This is under ____ search

Iterative deepening search, Uninformed

______ expands two frontiers, one around the initial state and one around the goal, stopping when the two frontiers meet

Bidirectional search, Uninformed

_______ expands nodes with minimal h(n). It is not optimal but is often efficient. This is under _____ search

Greedy best-first search, Informed

______ expands nodes with minimal f (n) = g(n)+h(n). This is under ____ search

A* search, Informed

______ is sometimes more efficient than A* itself. This is under ______ search

Bidirectional A* search, Informed

_____ is an iterative deepening version of A*, and thus addresses the space complexity issue. This is under ____ search

Iterative Deepening A* search, Informed

True or False: In a partially observable environment, Local search is the best way to do.

True

Hill Climbing and Simulated Annealing is an example of _____

Local Search

A/an ____ is a stochastic hill-climbing search in which a population of states is maintained.

Evolutionary Algorithm

In nondeterministic environments. agents can apply _____ search

And-or

When the environment is partially observable, the _______ represents the set of possible states that the agent might be in

Belief State

_____ arise when the agent has no idea about the states and actions of its environment

Exploration problems

_____ represent a state with a set of variable/value pairs and represent the conditions for a solution by a set of constraints on the variables

Constraint Satisfaction Problems (CSPs)

______ is a form of depth-first search, and is commonly used for solving CSPs

Backtracking Search

Game Theory is best applicable in multi agents that are _____

Competitive

This is how the board in a game is set up

Initial State

This says when the game is over

Terminal Test

In two-player, discrete, deterministic, turn-taking zero-sum games with perfect information, the ______ algorithm can select optimal moves by a depth-first enumeration of the game tree

Minimax

The ______ search algorithm computes the same optimal move as minimax, but achieves much greater efficiency by eliminating subtrees that are provably irrelevant

Alpha-beta

This tree search does not apply heuristic function, but by simulating all of its moves from start to end and choose the best move from it

Monte Carlo Tree Search

Stochastic games are just nondeterministic games with ____, such as dice, coin flipping, and many more

Probability

In a competitive environment, an agent must choose the state with the best ____

Utility

If the outcomes of all the states are equal, the agent must use ______

Nash Equilibrium

This is a table listing of all possible combinations of decision alternatives and states of nature (utility)

Payoffs table