Introduction to Artificial Intelligence

Exercises: https://www.ics.uci.edu/~welling/teaching/ICS171fall10/cs171_hw1_solutions.pdf

18 flashcards · Computer Science · by lakesare

Broadly speaking, AI is the study and design of computer systems that exhibit intelligence.
However, in practice there is a great deal of variation in what is considered intelligent, and how intelligence should be achieved. Here are 6 different definitions of the AI that reflect these differences: ’ ... effort to make computers think ... ’ (Haugeland, 1985)’[The automation of] activities that we associate with human thinking ...’ (Bellman, 1978)’The study of mental faculties through the use of computer models.’ (Charniak and McDermott, 1985)’The study of the computations that make it…
What is the goal of AI?
One goal, typified by Alan Turing’s famous ’Turing test’ for intelligence—in which a program is judged intelligent if its behaviour cannot be differentiated from that of a human - is to explicitly aim to replicate human behaviour. However, most modern AI programs are instead designed to act rationally, meaning that they take the best possible action given their goals, knowledge and constraints.
Most work in AI focuses on smaller components thought to be necessary for producing intelligent programs. The major subfields are:
1. Problem solving, where an agent is given a problem setting and a goal and must determine how to realize that goal. For example, given the axioms of number theory, an agent could try to prove that there are infinitely many prime numbers. 2. Knowledge representation and reasoning, which studies how an agent can represent knowledge it has about the environment and use it to derive further knowledge, either using a logic-based representation (when the knowledge is certain) or a probabilistic one…
An agent is anything that can be viewed as: perceiving its environment through sensors and acting upon that environment through actuators.
A human agent has eyes, ears, and other organs for sensors and hands, legs, vocal tract, and so on for actuators. A robotic agent might have cameras and infrared range finders for sensors and various motors for actuators.
What does the agent perceive environment though?
Sensors A human agent has eyes, ears, and other organs for sensors and hands, legs, vocal tract, and so on for actuators. A robotic agent might have cameras and infrared range finders for sensors and various motors for actuators.
What does the agent act upon environment through?
Actuators A human agent has eyes, ears, and other organs for sensors and hands, legs, vocal tract, and so on for actuators. A robotic agent might have cameras and infrared range finders for sensors and various motors for actuators.
What is meant by percepts?
It's what environment sends to agent's sensors. Agents are systems that interact with an environment using sensors to receive perceptual inputs (called percepts) from that environment, and actuators to act upon it. A web search agent has English search queries as percepts, accesses to a database of web pages and an output web page as actuators, and the Internet as an environment. The agent program queries the database to find the best set of matches to the search queries, and displays them on t…
The extent to which the agent uses prior knowledge (instilled in it by its designer) rather than its own experience, is the extent to which that agent lacks autonomy.
In general, we aim to build autonomous agents because they: require less expert knowledge, and are more adaptable than agents that require a great deal of problem-specific knowledge. However, autonomous agents are typically much harder to design—it is much easier to build a problem-specific agent with all the knowledge that it might need than to build a general one that can acquire that knowledge
What is intelligent agent?
An agent is anything that can be viewed as: perceiving its environment through sensors and acting upon that environment through actuators.
What is fully observable VS partially observable task?
fully observable task: each percept tells all information about the environment partially observable task: each percept only tells partial information, agent has to keep track If an agent’s sensors give it access to the PARTIALLY OBSERVABLE complete state of the environment at each point in time, then we say that the task environment is fully observable.A task environment is effectively fully observable if the sensors detect all aspects that are relevant to the choice of action; relevance, in t…
Name any 3 dimensions along which tasks may differ. We can consider a combination of performance measure, environment, sensors and actuators to be a task which the agent must solve.
Fully versus partially observable. In a fully observable task, the current percept reveals all of the relevant state of the environment, and therefore the agent does not need to internally keep track of the world. In a partially observable task the current percept only reveals some of the relevant information. Deterministic versus stochastic. In a deterministic task, the current state of the environment and the current action are sufficient to exactly predict the next state of the environment.…
Fill in empty spaces.
What's the complete history of everything the agent has ever perceived called?
Percept sequence
Define a rational agent.
For each possible percept sequence, a rational agent should select an action that is expected to maximize agent's performance measure, given the evidence provided by learning and whatever built-in knowledge the agent has.
In 'PEAS description of the task environment; - what does PEAS stand for?
P - Performance Measure E - Environment A - Actuators S - Sensors
Give an example of partially observable environment, and fully observable environment
partially observable - vacuum cleaner with only current-square sensor fully observable - chess
Name 4 basic kinds of agent programs.
Simple reflex agents - these agents select actions on the basis of the current percept, ignoring the rest of the percept history Model-based reflex agents - keeps an internal model of the world (state, how the world evolves, what my actions do). Goal-based agents - agent keeps both model of the world, and a goal - describes situations that are desirable. Utility-based agents - agents that keep model of the world, goal (binary success/failure, e.g. destination of taxi) AND utility (how happy age…
Symbolic VS sub-symbolic AI?
symbolic AI - uses mathematical symbols to represent objects and relationships between objects. sub-symbolic AI - neural networks On one side was an approach—now called “good old-fashioned AI”—that had dominated the field since the 1950s. Also known as symbolic AI, it used mathematical symbols to represent objects and the relationship between objects. Coupled with extensive knowledge bases built by humans, such systems proved to be impressively good at reasoning and reaching conclusions about d…