What is Intelligence?
The capacity of a system to compute, reason, recognize parallels and relationships, gain knowledge through experience, store and retrieve information from memory, solve problems, understand complex concepts, speak naturally, classify, and generalize.
The capacity to produce, communicate, and comprehend sound-based meanings; awareness of pitch and rhythm
Speaking, recognizing, and utilizing phonological (speech sounds), grammatical (grammar), and semantic (meaning) methods (meaning).
being able to perceive visual or spatial information, alter it, and recreate visual pictures without using the objects as a reference, as well as build 3D images and move and rotate them.
the capacity to use and comprehend relationships without action or external things. comprehending difficult and sophisticated concepts.
control over fine and coarse motor abilities, and the ability to manage things with the entire body or just a portion of it.
the capacity to distinguish between the intentions, beliefs, and feelings of another.
the capacity to differentiate between one's own thoughts, intents, and motives.
What Constitutes Intelligence?
Intelligence has no physical form. It is made up of
- Problem Solving
- Cognitive Linguistics
We are able to provide a basis for judgment, decision-making, and prediction thanks to a set of processes. Generally, there are two types:
1. It makes precise observations before making sweeping generalizations.
2. Inductive reasoning allows for the possibility of the conclusion being untrue even though all of the premises are true.
Example: "Nita is a professor. Nita is diligent. As a result, all teachers are diligent."
1. It begins with a broad assertion and considers all potential outcomes before coming to a clear, logical conclusion.
2. If something holds true for a group of objects in general, it holds true for all individuals who belong to that group.
It is the process of learning something new through study, practice, instruction, or experience. Learning increases knowledge of the study topics. Humans, some animals, and systems using AI are all capable of learning. Learning is classified as
Learning to behave in a given way in response to certain stimuli is known as stimulus-response learning. For instance, when the doorbell rings, a dog will lift its ear.
Learning by listening and hearing is known as auditory learning. Students could listen to lectures that have been recorded as audio.
Learning through the repetition of events that one has seen or experienced is known as episodic learning. This is logical and systematic.
Relational learning entails learning to distinguish between multiple stimuli based on relational characteristics rather than absolute characteristics. For instance, while cooking potatoes that were salty the prior time when you added, say, a tablespoon of salt, use "a bit less" salt this time.
Learning using visual cues like pictures, colors, maps, etc. is referred to as spatial learning. Before actually following the path, a person can, for instance, draw up a roadmap in their head.
Motor learning is learning through exact muscle movement. Picking up objects, writing, etc. are a few examples.
Observational learning is the process of learning by observing and copying others. The youngster, for instance, tries to learn by imitating her parent.
Learning to recognize stimuli that one has previously seen is known as perceptual learning. Identifying and categorizing things and circumstances, for instance.
Agent and environment: what are they?
An agent is something that has sensors for sensing its surroundings and effectors for acting on it.
1. A human agent has other organs, such as hands, legs, and the mouth, for effectors, in addition to sensory organs like the eyes, ears, nose, tongue, and skin, which are parallel to the sensors.
2. In place of numerous motors and actuators for the effectors and cameras and infrared range finders for the sensors, there is a robotic agent.
3. Encoded bit strings serve as the basis for a software agent's actions and programs.
Performance Measure of Agent
Agent Performance Metrics How successful an agent is is determined by the criteria.
Behavior of Agent
The agent's conduct is whatever the agent does in response to a particular pattern of perceptions.
The agent's perceptual inputs at a specific instance are known as percepts.
The history of everything an agent has ever perceived is known as the perceptual sequence.
A map from the precept sequence to action is the agent function.
What is Rationality?
1. The state of being logical, sensible, and intelligent is what is meant by being rational.
2. According to what the agent has perceived, rationality is concerned with anticipated behaviors and outcomes. An essential component of the reason is doing activities with the intention of learning meaningful information.
What does a rational ideal agent do?
One who is capable of doing predicted behaviors to optimize their performance measure is the ideal rational agent.
1. its perceptual pattern
2. Their own knowledge base
An agent's ability to reason depends on the following factors:
- The performance metrics define success levels.
- So far, the agent's perception sequence.
- preexisting environmental knowledge of the agent.
- The things an agent is capable of doing.
A rational agent always takes the proper action, where the proper action is defined as the action that will provide the agent the greatest chance of success in the specific perceptual sequence. The agent uses Performance Measure, Environment, Actuators, and Sensors to solve the issue (PEAS).
The Structure of Intelligent Agents
The structure of an agent can be characterized as
Agent = Architecture + Agent Program
Architecture = the equipment used to run agents.
An agent program carries out an agent function.
Simple Reflex Agents:
1. They only make decisions based on what they are now perceiving.
2. They can only be considered rational if the proper choice is made solely based on the accepted premise.
3. Their surroundings are entirely visible.
A state (condition) is mapped to an action by a rule known as the condition-action rule.
Reflex Agents Based on Models:
They make decisions about what to do based on a world model. They keep a state inside themselves.
They Built a model
Depending on the perceptual history, the internal state represents the current state's unobserved characteristics.
Updating the state requires knowledge of
1. the way the world changes.
2. What impact do the agent's deeds have on the world?
They decide what they will do in order to accomplish their objectives. Since the knowledge underpinning a choice is clearly modeled in a goal-based approach, it is more adaptable than a reflex agent.
Goal - the goal is a description of the ideal circumstances.
Agents Based on Services:
Based on a preference (utility) for each state, they decide which actions to do.
When goals are insufficient.
1. There are many competing objectives, but only a select number can be accomplished.
2. Goals have a chance of achievement, so you must balance their importance with the possibility of accomplishment.