Smart Systems Lec 1.pdf

# Smart Systems and Computational Intelligence ## Lecture 1: Smart Systems and Computational Intelligence Course - The first slide contains Arabic text with the logo of Alexandria University and Faculty of Computers and Data Science. - The lecture is titled "Smart Systems and Computational Intelli...

# Smart Systems and Computational Intelligence ## Lecture 1: Smart Systems and Computational Intelligence Course - The first slide contains Arabic text with the logo of Alexandria University and Faculty of Computers and Data Science. - The lecture is titled "Smart Systems and Computational Intelligence". - The lecturer is Mayar Mostafa, and their email address is [email protected]. ## Grading The course grade is determined by the following: | Category | Weight | |------------- |------:| | Year Work | 10 | | Final Project | 10 | | Practical | 10 | | Midterm | 20 | | Final Exam | 50 | ## Human Intelligence Human intelligence is a composition of different abilities, including: - Learning - Reasoning - Perceiving - Understanding of Language - Feeling ## What is Intelligence? Intelligence is the ability to acquire knowledge and use it. ## What is AI? AI is an artificial intelligence that can be defined as: - The study of how to make computers do things that, at the moment, people are better at (Elaine Rich) - The study of ideas that enable computers to be intelligent; that is, to reason, to apply knowledge, to perceive and manipulate things. - The part of computer science concerned with the design of computer systems that exhibit human intelligence. ## Definition of AI "Intelligence: The ability to learn and solve problems" (Webster's Dictionary). "Artificial Intelligence (AI): The intelligent exhibited by machines or software" (Wikipedia). "The science and engineering of making intelligent machines" (McCarthy). "The study and design of intelligent agents, where an intelligent agent is a system that perceives its environment and takes actions that maximize its chances of success," (Russel and Norvig AI book). ## Four Schools of Thoughts There are four schools of thoughts within the field of AI: | School of Thought | Description | |---|---| | Thinking Humanly | The exciting new effort to make computers think - machines with minds, (Haugeland, 1985). The automation of human thinking activities such as decision-making, (Hellman, 1978). | | Acting Humanly | The art of creating machines that perform functions, (Kurzweil, 1990). The study of how to make computers do things better, (Rich and Knight, 1991). | | Thinking Rationally | The study of mental faculties, (Charniak and McDermott, 1985). The study of computations that make it possible to perceive, reason, and act, (Winston, 1992). | | Acting Rationally | Computational Intelligence is the study of the design of intelligent agents, (Poole et al, 1998). AI... is concerned with intelligent behavior, (Nilsson, 1998). | ## The Turing Test - The Turing test is a test of a machine's ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human. - A human questioner interacts with a machine via a text-based interface. - The questioner tries to determine which terminal is operated by a human and which terminal is operated by the machine. - The computer is considered to have passed the Turing test if the questioner is unable to distinguish it from a human. ## Acting Humanly - The Turing Test Approach - The Turing test approach to AI focuses on creating machines that can pass the Turing test. - To pass the Turing test, a machine would need to possess the following capabilities: - Natural language processing to enable it to communicate successfully in English - Knowledge representation to store what it knows or hears - Automated reasoning to use the stored information to answer questions and to draw new conclusions - Machine learning to adapt to new circumstances and to detect and extrapolate patterns - Computer vision to perceive objects - Robotics to manipulate objects and move about ## Cognitive Computing vs. Artificial Intelligence | Category | Cognitive Computing | Artificial Intelligence | |------------- |------:|-----:| | Focus | Mimicking human behavior and reasoning | Augmenting human thinking | | Goal | Solve complex problems | Solve complex problems and provide accurate results | | Method | Simulates human thought processes | Finds patterns to learn or reveal hidden information | | Role | Supplements information for humans to make decisions | Makes decisions on its own | | Applications | Customer service, healthcare, industries | Finance, security, healthcare, retail, manufacturing | ## Thinking Rationally - Laws of Thought Approach - The laws of thought approach attempts to codify "right thinking" based on logic. - Greek philosopher Aristotle is one of the first to attempt to codify this approach. - The logicist tradition in AI hopes to create intelligent systems using logic programming. - Two obstacles to this approach: - It is not easy to take informal knowledge and state it in the formal terms required by logical notation. - Solving a problem principally is different from doing it in practice. ## Acting Rationally - The Rational Agent Approach - Rational agent approach is based on the idea that an intelligent system should act rationally - that is, in a way that maximizes its chances of success. - "Behave rationally" means taking the right action to achieve the goals based on the system's knowledge and belief. - The rational agent approach aims to create systems that can act rationally in a given environment. ## Goals of AI - To make computers more useful by letting them take over dangerous or tedious tasks from humans. - Understand the principles of human intelligence. ## Artificial Intelligence Applications - Gaming - Astronomy - Healthcare - Transport - Agriculture - Education - Data Security - Finance - Social Media - Automotive - Robotics - Entertainment - E-commerce ## Artificial Intelligence in Our Daily Life AI is used in many aspects of our daily lives: - Email (spam filtering) - Game playing - Robotics - Virtual assistants - ATM ## Intelligent Systems - An intelligent system is a machine designed to perform a task of some utility to humans. - The development of such a type of system is now an engineering discipline of information technology that requires effective methods and tools. - The precise characterization of an intelligent system is non-trivial because it is based on concepts related to cognition, an area that is not fully understood. ## Characterization Based on External Behavior - An intelligent system is understood as a machine designed to perform a task of some utility to humans. - An intelligent system is characterized by its external behavior, i.e., how it interacts with the environment. ### Agent-based operation - An intelligent system operates as an agent which means that it is able to act. - The agent senses the environment (e.g., through sensors) and acts in the environment (e.g., through actuators). - It interacts with other agents (e.g., human users or other artificial systems). ### Rational Behavior - Rational means the system performs actions that optimize a performance measure. - It can be analyzed from an outside perspective, without having to know the internal mechanisms that generate such behavior. - To act rationally, a machine should consistently take actions that successfully optimize the performance measure. - Acting rationally is related to intelligence. - For example, a chess player is considered smarter if they win more games. - The degree of intelligence is also determined by the diversity of tasks the agent is able to do. - An agent exhibits rational thinking if it is able to provide reasons for what it does or what it believes. An external observer can verify a system's rational thinking if it uses an understandable language to explain beliefs. ### Interacting with the Environment - An intelligent system operates as an agent, which means it is able to act using data about the environment. - The use of sensors and actuators separates the body of the intelligent system from the environment. This characteristic is called embodiment. - An agent is situated in an environment because it operates in a close-coupled interaction with the environment. - An intelligent system can interact with other agents (e.g., human users or artificial computer-based agents). - A simple example is a thermostat which senses the temperature using a thermometer and acts by starting or stopping a heater. The thermostat also communicates with a human user. #### Environment Properties The properties of environments include: | Property | Description | |---|---| | Static (or dynamic) | The environment does not change (or changes) while an agent is making a decision | | Discrete (or continuous) | The observed state of the environment, time or actions are discrete (or continuous) | | Fully-observable (or partially-observable) | Sensors detect (or do not detect) all aspects that are relevant to the choice of action | | Deterministic (or stochastic) | The next state of the environment is (or it is not) completely determined by the current state and the action | | Episodic (or sequential) | Actions do not have influence (or they have influence) future actions | | Known (or unknown) | The outcomes for all actions are known (or they are not known) by the agent in advance | - Typically, the information about the environment that an intelligent system needs to perform its task is not completely known or its complexity exceeds the information processing capacity of the intelligent system. - For example, sensors of autonomous robots usually obtain partial and noisy information from the environment. - For instance, the environment of a chess player is static, discrete, fully observable, deterministic, sequential and known. In the case of a self-driving car, the environment is continuous, partial observable, stochastic, sequential and known. ### Interacting with Other Agents - The human user is one type of agent with whom an intelligent system typically interacts. - An intelligent system can play two different roles when interacting with a user: - Delegate: the system acts in the environment. - Advisor: the user acts in the environment. - During interaction, the system may be proactive instead of passive. This means the system may take the initiative to perform a task based on its own goals. - The system can also interact with other agents that are part of a complex organization (e.g., other machines or other human users). ## Rational Behavior - It is said that an agent acts rationally if the decisions it makes about its actions seek to optimize a performance measure. - A financial analyst makes decisions about investments to maximize the economic profit. - A self-driving car selects the path to follow according to a performance measure that minimizes the travel time. - Rational action can be analyzed from an outside perspective. - An external observer may conclude a machine acts rationally if it consistently takes actions that optimize the performance measure. - Acting rationally is related to one of the meanings we typically associate to intelligence. - For example, a chess player is considered smarter if they win more games. - The degree of intelligence is also determined by the diversity of tasks the agent is able to do. ## Learning - Learning is one of the characteristics commonly associated with intelligent behavior. - Two forms of learning are distinguished: - Performance-based: A system learns if it improves its performance in carrying out tasks. - Knowledge acquisition: An agent learns by acquiring new beliefs. - An intelligent system can acquire new beliefs by: - Observation of the environment. - Reasoning from other beliefs. - Knowledge transfer between agents through the use of language. - Learning is also important for adaptation to changes in the environment. - Agents can acquire new beliefs from other agents. - This mechanism is an effective mechanism of social adaptation for groups of agents. ## Any Questions? ## Thanks This document highlights the foundations of smart systems and computational intelligence, exploring key concepts such as human intelligence, AI, rational behavior, and agent-based systems. The document explores the role of AI in various domains and discusses the implications for human-machine interaction.