CS620 Subjective Feb 2025 by DeepSeek

Caution · This document is made by · from lecture 125 to end on 2 feb 2025

For final term pptx ,highlighted reading material and file send msg on

This is only main types short definations to remember

1. Agent-Based Modeling (ABM)

Agent: An autonomous entity that can perceive its environment, make decisions, and act upon the environment. Environment: The space or context in which agents operate and interact. Interaction: The way agents communicate or influence each other and

Types of Agents

Mobile Agents: Agents that can move around the environment (e.g., turtles in NetLogo). Stationary Agents: Agents that cannot move (e.g., patches in NetLogo). Connecting Agents: Agents that link two or more agents (e.g., links in NetLogo).

Agents have properties (e.g., energy, speed) and behaviors (e.g., moving, eating). ABM is used to simulate complex systems by modeling individual agents and their interactions.

2. NetLogo Basics

Turtles: Mobile agents in NetLogo. Patches: Stationary agents that represent the environment. Links: Connecting agents that represent relationships between turtles.

Key Commands in NetLogo

ask: Used to command agents to perform actions. let: Used to define variables. ifelse: Used for conditional statements. with: Used to filter agents based on conditions. one-of: Randomly selects one agent from a group.

NetLogo is a popular tool for ABM

The world in NetLogo is divided into patches, and turtles move on these patches.

Links connect turtles to represent relationships (e.g., friendships, communication channels).

3. Agent Properties and Behaviors

Properties: Attributes of agents (e.g., energy, speed, color). Behaviors: Actions agents can perform (e.g., moving, eating, reproducing).

Types of Agent Properties

Global Properties: Properties shared by all agents (e.g., energy in a predator-prey model). Breed-Specific Properties: Properties unique to a specific breed of agents (e.g., wolves have fang-strength, sheep have wooliness).

Agents can have both shared and unique properties. Behaviors are defined using rules (e.g., if a wolf has low energy, it eats a sheep).

4. Agent Cognition

Cognition: The process by which agents make decisions based on their environment and internal state.

Types of Agent Cognition

Reflexive Agents: React to inputs using simple if-then rules (e.g., if there’s food, eat it).

Utility-Based Agents: Make decisions to maximize a utility function (e.g., maximize fuel efficiency in cars).

Goal-Based Agents: Have specific goals and take actions to achieve them (e.g., reach a destination).

Adaptive Agents: Learn from past experiences and change their behavior over time.

Reflexive agents are the simplest, while adaptive agents are the most complex.

Utility-based agents optimize a specific function (e.g., fuel efficiency).

Goal-based agents have clear objectives (e.g., reaching home or work).