3 of 26

Goals of AI

To make computers more useful by letting them take over dangerous or tedious tasks from human

Understand principles of human intelligence

AI can empower computers to handle tasks that pose a risk to humans or simply drain our productivity. For instance, in hazardous environments, robots can take over dangerous jobs, reducing the potential for accidents and injuries.

�Computers can handle repetitive and mundane tasks that tend to bore or exhaust us. By automating these processes, we can free up valuable time and mental energy for more stimulating and challenging endeavors. Imagine the possibilities: no more mind-numbing data entry, lengthy paperwork, or repetitive manual labor.��We have the power to design this revolution with an emphasis on collaboration between humans and machines. By working together, we can create a world where our skills are amplified by technology, leading to unprecedented progress and improved quality of life.�

4 of 26

The Main Topics in AI

Artificial intelligence can be considered under a number of headings:�

Search (includes Game Playing)

Representing Knowledge and Reasoning with it

Planning

Machine�Learning

Natural language processing (NLP)

Expert Systems

Interacting with the Environment �(e.g. Vision, Games, Robotics)�

Artificial Intelligence (AI) is an exciting and rapidly growing field that encompasses various aspects of computer science. It involves creating intelligent machines that can perform tasks that typically require human intelligence. AI can be considered under several fascinating headings, each focusing on different aspects of intelligence and problem-solving.

One aspect of AI is search, which includes game playing. This involves developing algorithms and techniques for computers to search through large sets of possible moves or actions to make informed decisions. From chess to video games, AI-powered systems can challenge human players and provide intelligent opponents.

Another aspect is representing knowledge and reasoning. AI aims to develop methods for computers to understand, represent, and reason with knowledge. By encoding information in a structured manner, machines can use logical reasoning to draw conclusions, solve complex problems, and make informed decisions.

Planning is another crucial area of AI. It involves designing algorithms that enable computers to generate sequences of actions to achieve specific goals. By creating efficient plans, AI systems can optimize processes and solve complex tasks with precision.

Learning is a fundamental aspect of AI. Machines can be trained to improve their performance based on experience and data. This field explores how computers can acquire knowledge, recognize patterns, and make predictions through various learning techniques such as supervised learning, unsupervised learning, and reinforcement learning.

Natural language processing (NLP) focuses on enabling computers to understand and interact with human language. This includes tasks such as speech recognition, machine translation, and sentiment analysis. AI-powered virtual assistants like Siri and Alexa are excellent examples of NLP applications.

Expert Systems aim to replicate the knowledge and decision-making capabilities of human experts in specific domains. By encoding expert knowledge into computer systems, AI can provide valuable insights and assistance in fields like medicine, finance, and engineering.

Interacting with the environment is a fascinating area of AI that involves enabling machines to perceive and understand the world around them. This includes computer vision, where machines can analyze and interpret images and videos, as well as robotics, where intelligent machines can perform physical tasks with precision.

As high school students, exploring the field of AI can open doors to a world of innovation, problem-solving, and limitless possibilities. Whether it's designing intelligent algorithms, developing cutting-edge applications, or understanding the ethical implications of AI, this field offers endless opportunities to shape the future. So, get ready to embark on an exciting journey into the world of Artificial Intelligence!

5 of 26

Search is the fundamental technique of AI.

Possible answers, decisions or courses of action are structured into an abstract space, which we then search.

We may want to search for the first answer that satisfies our goal, or we may want to keep searching until we find the best answer.

Search is indeed a fundamental technique in the field of AI, and it helps us find answers or make decisions in a structured and organized way. Imagine that you have a goal or a question in mind, and you need to find the best possible answer or solution.

To do this, we create an abstract space where all the possible answers or courses of action are represented. This space can be thought of as a vast collection of options, like pieces of a puzzle spread out in front of you.

When we talk about searching, it means that we explore this abstract space systematically to find the answer we're looking for. It's like looking through all the puzzle pieces to find the one that fits perfectly.

There are different ways we can approach search. Sometimes, we might want to find the very first answer that satisfies our goal. It's like finding the first puzzle piece that fits, even if it's not the best fit. Other times, we might want to keep searching until we find the absolute best answer or solution. It's like carefully examining each puzzle piece until we find the perfect fit.

6 of 26

How Google Search Works?

7 of 26

Knowledge Representation and Reasoning

Knowledge representation involves finding effective ways to represent information and knowledge within an AI system.

The goal is to capture the essential aspects of the real world or a specific domain in a format that can be understood and processed by the AI.

Components in Knowledge representation:

Logic
Rules
Semantic Net
Frame (representation of data)
Script

8 of 26

Examples of Knowledge Representation

Rules :formal logic rules to represent knowledge

Image Logic: DialogFlow, Google and Ibm Thomas J. Watson. “Knowledge Representation for Chatbot Design preliminary report.” (2020).

Rules : Google Calendar

Logic : formal logic rules to represent knowledge

Let's take the example of a virtual personal assistant, like Siri or Google Assistant, to illustrate the components of knowledge representation.

Logic:

Logic: Logic-based representations use formal logic rules to represent knowledge. These rules define relationships, conditions, and constraints that help AI systems make deductions and draw logical conclusions.

Suppose you ask your virtual assistant, "If it's raining outside, remind me to take an umbrella." The assistant uses logical rules to represent this knowledge. It understands that the condition "it's raining" should trigger the action "remind to take an umbrella." By applying the logical rule, it can remind you whenever it detects rain.

Rules:

: Rules are statements that specify conditions and actions. They provide a set of guidelines for reasoning and decision-making in AI systems. By applying these rules, AI can determine appropriate actions or draw conclusions based on the given conditions.

Your virtual assistant has a set of predefined rules. For example, it might have a rule that states, "If the user's calendar has an appointment, provide a reminder." This rule guides the assistant in determining when and how to provide timely reminders based on the user's schedule.

9 of 26

Semantic Net : represent knowledge using interconnected nodes and link

Frames :an object or concept

Image : devopedia.org

Image : Google search for restaurants near me

Semantic Net:

Semantic nets, or semantic networks, represent knowledge using interconnected nodes and links. Each node represents a concept, while links represent relationships between concepts. This graphical representation helps AI systems organize and understand the connections between different elements of knowledge.

The assistant maintains a semantic net to organize knowledge. Suppose you ask, "What is the capital of France?" The assistant uses its semantic net to link the concept of "France" with its associated capital, "Paris." It retrieves the relevant information from its knowledge base and provides you with the answer.

Frames:

Frames provide a structured way to represent knowledge by organizing it into a collection of attributes and values associated with an object or concept. Each frame contains information about specific characteristics, properties, or behaviors related to the object or concept.

The virtual assistant uses frames to store information about specific entities or concepts. For instance, it may have a frame for "restaurants," which contains attributes like "name," "cuisine," "location," and "reviews." When you ask for restaurant recommendations, the assistant accesses these frames to provide relevant information based on your preferences.

10 of 26

Scripts:

Scripts capture knowledge about events or sequences of actions.

Image : Google Blog, Two new ways the Google Assistant can help with an upcoming trip

Scripts:

Scripts capture knowledge about events or sequences of actions. They describe the typical steps, roles, and outcomes of specific activities. Scripts help AI systems understand and predict the expected behavior in a given context, enabling them to make informed decisions or generate appropriate responses.

Your virtual assistant has scripts that capture knowledge about common sequences of actions. For example, it knows the steps involved in booking a flight, such as searching for available flights, selecting a suitable option, providing passenger details, and making the payment. By following the script, the assistant can guide you through the flight booking process

In this way, the components of knowledge representation work together in the virtual assistant system. Logic and rules enable decision-making, semantic nets organize knowledge relationships, frames store specific information, and scripts guide the assistant's behavior in various tasks.

11 of 26

Planning

Planning in AI refers to the process of generating a sequence of actions or a strategy to achieve a specific goal. It involves determining the optimal sequence of actions that an AI agent should take, considering the current state of the environment, the desired goal, and any constraints or limitations.

Planning in AI is used in various applications, such as robotics, autonomous vehicles, logistics, scheduling, and game playing. It enables AI systems to make informed decisions, solve complex problems, and achieve goals in dynamic and uncertain environments

12 of 26

Planning Applications

Autonomous Planning & Scheduling:

Autonomous rovers.
Telescope scheduling
Transportation:

Autonomous vehicle control

13 of 26

Amazon Warehouse

14 of 26

Expert Systems

Experts Systems designed to mimic the decision-making abilities of human experts in a particular domain. They are computer programs that incorporate knowledge and rules provided by human experts to solve complex problems or provide expert-level advice.

Expert systems are widely used in various fields, including healthcare, finance, engineering, and troubleshooting. They excel at solving complex problems, providing consistent and reliable advice, and capturing and preserving expert knowledge that may be difficult to access or retain.

15 of 26

Expert Applications

Medicine:

Image guided surgery

16 of 26

Natural Language Processing (NLP)

NLP is a subfield of Artificial Intelligence (AI) that focuses on enabling computers to understand, interpret, and generate human language in a way that is both meaningful and useful. NLP involves the development of algorithms, models, and techniques to process, analyze, and generate natural language data.

17 of 26

Medical Natural Language Processing

18 of 26

Interacting with the Environment

Interacting with the environment often involves a feedback loop. AI systems receive feedback in the form of new observations or outcomes resulting from their actions. This feedback helps them update their understanding of the environment, refine their decision-making process, and learn from experience. Techniques such as reinforcement learning, where the AI system learns from rewards or penalties, are commonly used for this purpose.

19 of 26

Computer Vision Application

Transportation:

Pedestrian detection

20 of 26

Computer Vison Applications

Image analysis and enhancement

21 of 26

How Computer Vision Works?

23 of 26

Excercise

Quick, Draw! is an online game developed by Google that challenges players to draw a picture of an object or idea and then uses a neural network artificial intelligence to guess what the drawings represent. The AI learns from each drawing, improving its ability to guess correctly in the future.��Go to following link: ��https://quickdraw.withgoogle.com/

WiFi Name : AIISC_FRONT_CONF/ AIISC_FRONT_CONF5G

PASSWORD: AIISC@2022

1 of 26

2 of 26