REVIEWER-CS-ELEC-3.pdf

CS ELEC 3- INTELLIGENT SYSTEM o The Turing Test (developed by Alan Turing) is used to measure a machine’s abili...

CS ELEC 3- INTELLIGENT SYSTEM o The Turing Test (developed by Alan Turing) is used to measure a machine’s ability to exhibit intelligent behavior Lecture 1 to 5. An Overview of Artificial Intelligence and equivalent to, or indistinguishable from, Key Concepts that of a human. This lecture will guide you through the main topics that o Recent advancements have focused on are essential to understanding AI. machine learning, deep learning, and natural language processing (NLP). 1. What is Artificial Intelligence? 3. Types of AI and Examples Artificial Intelligence refers to the capability of a machine to imitate intelligent human behavior. The primary goal of Narrow AI: Focused on one specific task. AI is to create systems that can perform tasks that o Examples: Email filtering, search engines, normally require human intelligence, such as reasoning, and self-driving cars. learning, problem-solving, and understanding language. General AI: Has the potential to perform any Primary Goal of AI: cognitive task that a human can. o To create machines that can replicate o This remains a theoretical concept today. human intelligence. Superintelligent AI: Hypothetical form of AI that o AI is not just about creating robots or could outperform humans in all domains. software for specific tasks; it involves developing machines that can solve 4. The Role of Machine Learning in AI problems autonomously and adapt over Machine learning (ML) is a subset of AI that enables time. machines to learn from data and improve over time Key Concepts: without being explicitly programmed. Narrow AI: Systems designed for a specific task, Supervised Learning: Involves learning from such as a virtual assistant like Siri or Alexa. labeled data. General AI: Systems that can perform any Unsupervised Learning: Involves discovering intellectual task that a human can. patterns in unlabeled data. Superintelligent AI: Hypothetical AI that Reinforcement Learning: Focuses on training surpasses human intelligence in all aspects. models through a system of rewards and penalties. 2. History and Evolution of AI Key Difference: The field of AI officially started in 1956 at the Dartmouth Conference, where researchers defined AI as a field of Supervised Learning uses labeled data for study. This event marked the birth of AI as a separate training, while Unsupervised Learning discovers discipline. hidden patterns in data without predefined labels. Milestones: 5. Knowledge Representation in AI o Early systems like Logic Theorist and General Problem Solver focused on Knowledge representation is crucial for AI to store problem-solving. information in a way that enables reasoning.  Logic-based Representation: Uses propositional Integration: AI enables robots to learn, perceive their and predicate logic. environment, and make autonomous decisions. Semantic Networks: Represent concepts using Logic is the study of principles of correct reasoning. It nodes and relationships. helps us understand what follows from what, ensuring that our conclusions are valid based on given premises. Frames: Data structures for representing stereotyped situations. Key Concepts in Logic Importance: 1. Principles of Reasoning: Enables machines to understand, reason, and o Logic is concerned with the principles make decisions based on complex data. of correct reasoning. 6. Challenges and Ethical Considerations in AI o It ensures that conclusions follow logically from premises. AI faces several challenges, including: 2. Process of Logic: Bias in AI algorithms: AI systems can inherit biases from training data. o Inputs in logic are called premises. Ethical concerns: Includes decisions about o Outputs are called conclusions. privacy, security, and the long-term impact of AI. 3. Symbolic Logic: Job displacement: Automation may lead to o A method of representing logical unemployment in certain sectors. statements using symbols and variables. Ethical AI Development: o Helps in simplifying and solving logical It is crucial to ensure that AI is used for good problems. purposes and does not pose a threat to society. Propositional Logic 7. AI in Addressing Global Challenges 1. Basic Concepts: AI can be a powerful tool for solving global issues such as: o Deals with statements that can be Healthcare: Diagnosing diseases and either true or false. personalizing treatments. o Well-formed atomic propositions are Climate Change: Predicting environmental represented by letters like A, B, C, or changes and optimizing resource usage. with subscripted numerals. Education: Creating personalized learning 2. Manipulating Logical Variables: experiences. o Focuses on manipulating logical variables that represent propositions. 8. The Relationship between AI and Robotics o Uses Boolean algebra for efficient reasoning. AI and robotics are often intertwined but are distinct fields: 3. Advantages: Robotics: Focuses on building physical machines. o The Boolean nature of propositional logic allows for efficient and straightforward AI: Concerned with creating intelligent behavior reasoning. in machines. First-Order Predicate Logic (FOPL) o Represents the logical AND operation. 1. Extended Expressiveness: 3. Disjunction (OR): o Extends propositional logic by allowing o Symbol: ∨ reasoning about objects and their o Represents the logical OR operation. relationships. o Uses variables and predicates to make more complex statements. Expert System/ Robotics 2. Predicates and Functions: Expert systems are a type of artificial intelligence (AI) designed to emulate the decision-making abilities of a o Predicates in FOPL are similar human expert. When applied to robotics, expert systems to Boolean functions in programming. can significantly enhance a robot’s capabilities in various o They define a relation between two ways: atoms. 1. Decision-Making: Expert systems can help robots 3. Constants and Variables: make decisions based on a set of rules and knowledge stored in their database. This allows o Constants are often represented robots to handle complex tasks and make by numbers or names. informed decisions in real-time1. o Variables stand for unspecified quantities. 2. Problem-Solving: By using a knowledge base and 4. Functions and Terms: inference engine, expert systems enable robots to solve problems that require human-like o Functions represent non-Boolean values. reasoning. This is particularly useful in o Arguments to predicates and functions environments where robots need to adapt to new can only be terms. situations2. Quantifiers in FOPL 3. Learning and Adaptation: Expert systems can be designed to learn from their environment and 1. Universal Quantifier: experiences. This means that robots can improve o Expresses “for all” or “every”. their performance over time by updating their knowledge base with new information2. o Symbol: ∀ 4. Control and Coordination: In multi-robot systems, 2. Existential Quantifier: expert systems can coordinate the actions of o Expresses “there exists” or “some”. multiple robots, ensuring they work together efficiently. This is essential in applications like o Symbol: ∃ manufacturing, where precise coordination is Logical Symbols required3. 1. Negation: 5. Human-Robot Interaction: Expert systems can enhance the interaction between humans and o Symbol: ¬ (or ~) robots by enabling robots to understand and o Represents the negation of a statement. respond to human commands more effectively. This makes robots more intuitive and 2. Conjunction (AND): easier to work with1. o Symbol: ∧ (or &) Overall, integrating expert systems into robotics allows for What it does: Helps robots see and recognize more intelligent, adaptable, and efficient robotic behavior, objects. making them capable of performing tasks that were Example: Like facial recognition on your phone. previously thought to be too complex for machines. 4. Autonomous Navigation: What it does: Helps robots move around without human help. Role of Algorithms in Robotic Tasks Example: Like drones flying on their own. 1. Path Planning: 5. Human-Robot Interaction: What it does: Helps robots figure out the best What it does: Makes robots better at route to take to reach a destination. understanding and responding to humans. Example: Like using a map to find the shortest Example: Like voice assistants that understand way to your friend’s house. your commands. 2. Obstacle Avoidance: These technologies make robots smarter and more What it does: Helps robots detect and avoid adaptable, allowing them to perform complex tasks more obstacles in their path. efficiently. If you have any specific questions or need more details, feel free to ask! Example: Like a car’s sensors that stop it from hitting something. 3. Manipulation: Artificial Intelligence (AI) encompasses a wide range of technologies designed to perform tasks that typically What it does: Controls robotic arms to pick up, require human intelligence. move, and place objects. 1. Medical Expert System: Example: Like a robot in a factory assembling parts. o Medical expert systems use AI to diagnose diseases by analyzing medical Enhancing Robots with Machine Learning data and symptoms, providing accurate 1. Learning from Data: diagnoses without the need for traditional tools. What it does: Robots analyze data to improve their tasks. 2. Virtual Assistant or Chatbot: Example: Like a robot vacuum learning the layout o Virtual assistants and chatbots use of your house to clean better. natural language processing (NLP) to understand and respond to user queries, 2. Real-Time Decision Making: helping with tasks like scheduling, What it does: Robots make quick decisions based information retrieval, and more. on current information. 3. Surgical Robot: Example: Like a self-driving car deciding when to o Surgical robots assist in performing stop or go. precise and minimally invasive surgeries, 3. Improved Perception: enhancing the capabilities of human surgeons. 4. Data Mining Expert System: scenes, and activities in images and videos. o These systems use algorithms to sift through large datasets, uncovering 12. Mobile Robot: patterns and insights that can inform o Mobile robots use sensors and algorithms decision-making and predictions. to navigate and avoid obstacles, enabling 5. Inference Engine: them to move autonomously in various environments. o Inference engines are a core component of expert systems, using logical rules to 13. Robot Controller: derive conclusions from known facts. o Robot controllers manage the precise 6. Symbolic Logic: movements of robotic arms, ensuring accurate and efficient task execution. o Symbolic logic uses symbols to represent logical statements, making it easier to 14. Affective Computing System: manipulate and solve logical problems. o Affective computing systems analyze 7. Machine Learning System: emotional cues from humans and respond appropriately, enhancing human- o Machine learning systems analyze data to computer interaction. learn and adapt, improving their performance on tasks through 15. Machine Learning System: experience. o Similar to the previous machine learning 8. Natural Language Processing System: system, these systems continuously learn and adapt to improve their performance. o NLP systems enable computers to understand, interpret, and respond to 16. Artificial Intelligence System: human language, facilitating o AI systems encompass a broad range of communication between humans and technologies designed to perform tasks machines. that typically require human intelligence. 9. Remote-Operated Robot: 17. Robotic System: o These robots are controlled remotely to o Robotic systems automate repetitive or perform tasks in hazardous environments, dangerous tasks, improving efficiency and such as deep-sea exploration or bomb safety in various industries. disposal. 18. Data Mining System: 10. Expert System: o Data mining systems extract valuable o Expert systems use a knowledge base and insights from large datasets, aiding in inference rules to provide advice and decision-making and strategic planning. solutions in specific domains, such as medical diagnosis or financial planning. 19. Virtual Reality System: 11. Computer Vision System: o Virtual reality systems create immersive digital environments for training, o Computer vision systems analyze visual entertainment, and education. data to identify and interpret objects, 20. Fraud Detection System:  o These systems analyze transaction data to o NLP systems enable computers to identify and prevent fraudulent activities. understand and interact with human language, facilitating communication and 21. Recommendation System: information retrieval. o Recommendation systems analyze user 30. Expert System: preferences and behavior to suggest products, services, or content. o Expert systems use a knowledge base and inference rules to provide expert advice 22. Control System: and decision-making support. o Control systems manage and regulate the 31. Robotics System: behavior of other systems, ensuring optimal performance. o Robotics systems control the movement and actions of physical robots, enabling 23. Virtual Human: them to perform various tasks. o Virtual humans are digital representations 32. Creative AI System: of people, used in simulations, games, and virtual environments. o Creative AI systems generate original content, such as art, music, and writing, 24. Self-Driving Car System: using AI algorithms. o These systems enable cars to navigate 33. Assistive Technology System: and drive autonomously, using sensors and AI to make driving decisions. o Assistive technology systems provide support and assistance to individuals with 25. Intelligent Agent System: disabilities, enhancing their quality of life. o Intelligent agents are autonomous entities that perceive their environment and take actions to achieve specific goals. 26. Virtual Assistant System: o Virtual assistant systems use AI to assist users with tasks like scheduling, reminders, and information retrieval. 27. Robotics System: o Robotics systems are designed to operate in hazardous conditions, performing tasks that are too dangerous for humans. 28. Machine Learning System: o Machine learning systems continuously GRF, 2024 learn from data and adapt to new scenarios, improving their performance over time. 29. Natural Language Processing System: