Understanding Artificial Intelligence: Definitions

Artificial Intelligence Definition

• AI is a machine that simulates human thought, takes actions based on those thoughts, and draws its own conclusions.
• It is capable of correcting itself.
• AI-based computer makes decisions in situations like humans, and sometimes even better.

Niti Aayog Definition

• AI is the ability of machines to perform cognitive tasks such as thinking, perceiving, learning, problem-solving, and decision-making.
• It has surpassed its initial conception by collecting and processing data.
• AI systems can handle tasks, facilitate connectivity, and improve productivity.

Common AI Applications

• Virtual Assistants: Assistants like Amazon's Alexa, Apple's Siri, Google Assistant, and Microsoft's Cortana use AI to understand voice commands, perform tasks, answer questions, and provide personalized recommendations.
• Recommendation Systems use AI algorithms to analyze behavior and preferences, and recommend content, products, or services tailored to individual tastes on platforms like Netflix, Spotify, and Amazon.
• Social Media platforms like Facebook, Twitter, and Instagram use AI algorithms for content moderation, personalized content curation, targeted advertising, and sentiment analysis.
• Navigation Apps like Google Maps and Waze utilize AI to analyze real-time traffic, provide optimal routes, estimate arrival times, and offer alternative routes to avoid congestion.
• Online Shopping: AI is used on e-commerce platforms such as Amazon and Alibaba for product recommendations, dynamic pricing, fraud detection, inventory management, and customer service chatbots.
• Healthcare uses AI for medical imaging analysis, disease diagnosis, drug discovery, personalized treatment recommendations, and remote patient monitoring.
• Finance uses AI for fraud detection, algorithmic trading, credit scoring, risk assessment, customer service chatbots, and personalized financial advice.
• Smart Home Devices: AI allows smart home devices like thermostats, security cameras, doorbell cameras, and lighting systems to automate, increase energy efficiency, and enable remote control via voice commands or mobile apps.
• Autonomous Vehicles use AI to perceive their surroundings, navigate safely, and make real-time driving decisions using sensor data and machine learning algorithms.
• Language Translation: AI-driven language translation services like Google Translate and Microsoft Translator use neural machine translation techniques to accurately translate text and speech between multiple languages.

Advantages of AI

• Automation: AI automates repetitive tasks, freeing human resources for creative and strategic activities, leading to increased productivity and efficiency.
• Accuracy and Precision: AI systems process large amounts of data with accuracy and precision, minimizing errors and improving decision-making quality.
• 24/7 Availability: AI-powered systems can operate continuously for 24/7 availability of services, such as customer support, virtual assistants, and online transactions.
• Personalization: AI algorithms analyze user data and behavior to provide personalized recommendations, content, and services for better user experience.
• Predictive Analytics: AI enables predictive analytics to forecast trends, anticipate customer needs, and identify potential risks or opportunities for better decision-making.
• Efficient Resource Utilization: AI-driven optimization algorithms optimize resource allocation, scheduling, and logistics for cost savings, reduced waste, and improved resource utilization.
• Enhanced Healthcare: AI technologies transform healthcare by assisting in medical imaging analysis, disease diagnosis, drug discovery, personalized treatment planning, and remote patient monitoring.
• Improved Safety: AI-powered systems enhance safety by detecting anomalies, predicting hazards, and implementing preventive measures in various domains.
• Innovative Applications: AI enables the development of new products, services, and business models with applications like natural language processing, robotics, virtual reality, and augmented reality.
• Global Competitiveness: AI gives organizations an edge in the global marketplace by leveraging advanced technologies to drive innovation, improve efficiency, and deliver products and services.

Disadvantages of AI

• Job Displacement: Automation may replace human workers in repetitive tasks, leading to job displacement and unemployment, especially in manufacturing, customer service, and administrative roles.
• Bias and Discrimination: AI algorithms can perpetuate biases in training data, leading to discriminatory outcomes related to race, gender, ethnicity, and more.
• Privacy Concerns: AI technologies that rely on large amounts of personal data raise concerns about privacy and data security, particularly from unauthorized access, data breaches, and misuse.
• Lack of Transparency: AI models, like deep learning algorithms, can be opaque and difficult to interpret, leading to a lack of transparency in decision-making processes.
• Overreliance on Technology: Overreliance on AI systems without human oversight can lead to overconfidence and complacency, especially in safety-critical domains like autonomous vehicles and healthcare.
• Ethical Dilemmas: AI raises ethical dilemmas such as the use of autonomous weapons, surveillance, and predictive policing, leading to concerns about accountability, human rights, and erosion of civil liberties.
• Social Isolation: Proliferation of AI technologies, like virtual assistants and social robots, may contribute to social isolation by reducing face-to-face interactions.

AI, ML & DL

• AI encompasses any technique enabling computers to mimic human intelligence, such as facial recognition and voice command understanding.
• ML(Machine Learning) is a subset of AI that enables machines to improve at tasks with experience by learning from data.
• DL(Deep Learning) trains software to perform tasks with vast amounts of data, making machines develop algorithms independently and efficiently.
• Deep Learning is the most advanced form of AI, followed by Machine Learning, while Artificial Intelligence covers different concepts and algorithms that mimic human intelligence.

Types of AI

• Narrow AI (Weak AI) is designed for specific tasks but lacks general intelligence, examples being Siri, Alexa, recommendation systems, image recognition, and language translation tools.
• General AI (Strong AI or AGI): Systems possess human-like intelligence with understanding, learning, and reasoning capabilities to perform any intellectual task, remaining theoretical.
• Super AI surpasses human intelligence and capabilities, remaining a distant goal with ethical and societal concerns.

AI Domains

• Data Science is related to data systems and processes, involving data collection, maintenance, and derivation of meaning.
• Computer Vision focuses on enabling machines to interpret and make decisions based on visual data. Examples are self-driving cars and face lock on smartphones.
• Natural Language Processing (NLP) Deals with the interaction between computers and humans using natural language to extract from spoken or written words. Examples are email filters and smart assistants.

AI ethics

• AI ethics involves moral principles, guidelines, and standards governing the development, deployment, and use of AI technologies to ensure fairness, responsibility, and benefit.
• Fairness and Bias: Ensuring AI systems are fair and unbiased prevents discrimination and ensures equal treatment across demographic groups.
• Transparency and Explainability: AI systems should be transparent and explainable, enabling users to understand the reasoning behind decisions.
• Privacy and Data Protection: Protecting individuals' privacy and data rights is essential, adhering to regulations, anonymizing data, and getting informed consent.
• Accountability and Responsibility: Establishing clear accountability is necessary to address ethical implications and consequences, so developers, organizations, and policymakers should take responsibility.
• Safety and Security: Ensuring the safety and security of AI systems is paramount to prevent accidents, errors, and malicious exploitation.
• Human-Centered Design: Prioritizing human well-being, autonomy, and dignity is essential to enhance societal benefit and minimize harm.
• Social Impact and Equity: Assessing and mitigating the social impact of AI helps to promote inclusive and equitable outcomes by addressing issues such as job displacement and inequality.
• Global Collaboration and Governance: Promoting international collaboration and cooperation ensures alignment with shared ethical principles and values.
• Another aspect to AI Ethics is bias and how any bias can transfer from the developer to the machine while the algorithm is being developed.
• As Artificial Intelligence is still a developing technology, not everyone can access Al enabled devices, leaving others behind.

AI project cycle

• Problem Scoping: Identifying a problem and having a vision to solve it.
• Data Acquisition: Collecting and gathering data from various sources for the AI project. This can be a piece of information or facts and statistics collected together for reference or analysis.
• Data exploration: Examining, cleaning, and visualizing data to discover patterns, trends, and relationships.
• Modelling: Creating mathematical representations of real-world phenomena or patterns in data using ML or deep learning techniques.
• Evaluation: Assessing the performance of the model to generalize new, unseen data, and whether it meets the project's success criteria.
• Deployment: Putting the trained and validated model into a production environment where the machine can make real-time predictions or decisions based on new data.

Sustainable Development Goals (SDGs)

• The Sustainable Development Goals (SDGs) are a collection of 17 global goals set by the United Nations General Assembly in 2015 as part of the 2030 Agenda for Sustainable Development.
• These are a range of interconnected global challenges, including poverty, inequality, environmental degradation, and climate change.
• Artificial intelligence can play a significant role in advancing the Sustainable Development Goals (SDGs).

AI's contributions towards SDGs

• Al can help governments and organizations better target poverty reduction efforts by analyzing socioeconomic data, identifying vulnerable populations, and optimizing resource allocation to achieve no poverty.
• Al-driven precision agriculture techniques can improve crop yield prediction, optimize resource use, and detect pests and diseases early, helping to ensure food security and reduce food waste.
• AI can improve healthcare delivery and outcomes through applications in healthcare, such as medical imaging analysis, predictive analytics for disease diagnosis and outbreak detection, and personalized treatment recommendations
• Al-based educational technologies, such as personalized learning platforms, intelligent tutoring systems, and language translation tools, promote quality education.
• Al can help reduce inequalities by improving disparities in access to resources and opportunities by analyzing data, informing policy decisions, and designing interventions to reduce inequality within and among countries.
• Al can optimize cities with planning-driven urban, transportation optimization, waste management, and public service delivery, which provide sustainable, resilient, and inclusive cities and communities supporting responsible consumption and production.
• Al can promote climate action by enhancing climate modeling, supporting renewable energy integration, optimizing resource management, and facilitating climate adaptation and mitigation efforts
• Al can protect life below water by helping marine ecosystems, combating illegal fishing, and promoting sustainable fisheries management.
• Al can aid in monitoring and managing terrestrial ecosystems, combating deforestation and desertification, protecting biodiversity, and promoting sustainable land use practices.
• Al can support conflict prevention, peacebuilding, and the rule of law by analyzing social media data for early warning signs of conflict, facilitating access to justice, enhancing peace, justice, and strong institutions.
• Al can facilitate collaboration and knowledge sharing among stakeholders and strengthen partnerships for achieving the SDGs.

4Ws problem canvas

• The “Who” block helps in analysing the people getting affected directly or indirectly due to it and finding who the ‘Stakeholders' to this problem are and what we know about them.
• Under the “What” block, determine the nature of the problem and explain how you know that there is a problem.
• Focus on the context/situation/location of the problem and look into the situation in which the problem arises for the "Where."
• In the “Why” canvas, think about the benefits which the stakeholders would get from the solution and how it will benefit them as well as the society.

Data Acquisition

• Data acquisition is the process of collecting and gathering data from various sources.
• Types of data
  • Training Data refers to the subset of data used to train machine learning models. Testing data, also known as test data or validation data, is a separate subset of data that is used to evaluate the performance and generalization ability of machine learning models after they have been trained on the training data.
• Ways to collect data:
  • Surveys and questionnaires
  • Interviews
  • Observational Studies
  • Experiments
  • Data Mining
  • Web Scraping
  • Sensor Data Collection
  • Focus Groups

Data Exploration

• A system map shows the components and boundaries of a system and the components of the environment at a specific point in time.
• Data exploration is extracting insights from the underlying structure of the data before analyzing it.

Data visualization

• Data visualization helps to make complex data more understandable.
• Common types of Data Visualization
  • Bar Charts: Useful for comparing quantities across different categories.
  • Line Charts: Ideal for displaying data trends over time.
  • Pie Charts: Show the proportional contributions of different categories to a whole.
  • Histograms: Display the distribution of numerical data by showing the frequency of data within certain ranges or bins.
  • Heat maps: Use color to represent data values in a matrix format.
  • Box Plots: Summarize data distribution through quartiles and highlight outliers.
  • Bubble Charts: Show relationships between three variables.
  • Treemaps: Making comparisons between different categories.
  • Gantt Charts: Used for project management to show the timeline of tasks or activities. Geographical Maps: Useful for visualizing data that has a spatial component.

Modelling

• Modelling refers to creating mathematical representations of real-world phenomena.
• Artificial intelligence models apply different algorithms to relevant data inputs to achieve the tasks, or output, they've been programmed for.
• AI models are generally classified as : Rule Based Approach and Learning Based Approach.
• Rule-Based Approach: In Al modelling, where the rules are defined by the developer, the machine follows the rules or instructions mentioned by the developer and performs its task accordingly.
• Learning-Based Approach: In Al modelling, the machine learns by itself. Under the Learning Based approach, the Al model is trained on the data fed to it and designs a model adaptive to the change in data.
• Types of learning models: Supervised learning, classification, and unsupervised learning

Supervised learning

 - The dataset which is fed to the machine has data that is labeled. The dataset is known to the person who is training the machine. A label is information used as a tag for data.

Classification

- A fundamental task where the goal is to categorize input data into several predefined classes or categories.

Regression

 - This focuses on predicting continuous numerical values.

unsupervised Learning

- The data which is fed to the machine is random and there is a possibility that the person who is training the model does not have any information regarding it.
 - Used to identify relationships, patterns, and trends out of the data which is fed into it.

unsupervised Learning Categories

• Clustering refers to the unsupervised learning algorithm that clusters the unknown data patterns.
• Dimensionality Reduction helps to reduces dimensions because there are various entities that exist beyond 3-Dimensions.

Evaluation

• Evaluation helps assess the model's performance to understand how well it can generalize to new data and whether it meets the project's success criteria.

Deployment

• Deployment is where the trained and validated model is put into a production environment where real-time predictions or decisions can be made based on new data.

Neural network’s

• Neural networks, also referred to as artificial neural networks (ANNs), are a class of machine learning algorithms inspired by the structure and function of the human brain.
  • Neurons, Units, Weights, etc.
    • Neurons (Nodes) are the basic computational units.
    • Weights and Biases determine the strength of a connection.
    • Activation Function: which determines its output, including the sigmoid, tanh, ReLU, and softmax functions.

Neural networks features

• Neural network systems are modelled on the human brain and nervous system.
• They extract features automatically without feeding the input by the programmer.
• Every node of layer in a Neural Network is compulsorily a machine learning algorithm.
• Very useful to implement when solving problems for big datasets.

Python

• Python is a high-level programming language, developed by Guido Van Rossum in 1991.
  • is free to use, and an interpreted & portable language, with a syntax that has a wide range of built-in functions, modules & libraries.

Python Installation

• Is available on www.python.org, which carry the python installation package.

Working Modes of Python

• Interactive Mode: User gives one command at a time and python executes the command and produces the output.
• Script Mode: In this mode, multiple statements are written and saved as a file in .py extension and executed.

Python Fundamentals

• The fundamental and smallest unit of a program is a token.
• There are five categories of tokens: Keywords, Identifiers, Literals, Operators, & Punctuators

Keywords

• Keywords are reserved words with a special meaning that cannot be identifiers.

Identifiers

• These are names given to different parts of program like variables, objects, classes, functions, and cannot start as digits or be a reserved keyword.

Literals and Values

• Literals are data items that have a fixed value.
• Python supports several kinds of literals: String literal, Numeric literals, Boolean Literals, Special Literals, & Literal Collections.
• The numeric literals in Python can belong to any of the following numerical types:
  • (1) Integer Literals
  • (2) Floating point Literals: Real literals are numbers having fractional values.

Boolean Literals

• Represent one of the two-boolean values, True or False.
• These are the only two values supported for Boolean Literals

Special Literals

• Python has one special character with no value, known as None.

Operators

• Operators perform specific operations when acted on variables, ex: a+b ; a & b , ‘+’ is the operator.
• Operators may be Unary or Binary

Conditional Operators

• It compares the values, such as greater than, less than, equal to, no equal, etc

Logical Operators

• Include (and, or, not)

Numbers

• Number data types are used to store numeric values.

Assignment Operators

• Adds values based on whether the left-side operands value is assigned, or the right-side operand of a designated value, etc

Punctuators and Delimiters

• Help implement the grammatical and structural syntax.

Input and output in Python

• Input () function in python is used to allow user to give input values via keyboard in form of string.
• In python we use Int() and float ()functions to convert the string values in integer and float respectively.

Data conversion

• Python allows for converting one data over to another with type conversion if it is done by a programmer, vs type-casting if done automatically by the compiler.

Print function

• The print() function in python is used to give output to user.

Variable

• A variable is a location or container that holds some value.
• Python is dynamic, therefore the type is automatically assigned and data type of variable can be changed.

Datatypes

• Specifies the type of data that we’re going to store in any variable
• Python support following data types with examples: - Numbers (int, float, complex, etc) - Strings (“Test”) - List [“test, “next”, “pass”] - Tuple (“test”, “2024”) - Dictionary {“test”: “string”, “function”}

Python loops

• In python are the “for”, “while” loops.
• The "for" loop will iterate until user tells it to stop or “break” the loop
• The "While" loop will continue as long as the loop is confirmed to be ‘True’.