1: Why AI?
2: What is Artificial Intelligence?
3: Types of AI
4: Key Branches of AI
5: How Machines Learn: Intro to Machine Learning
6: Deep Learning and Neural Networks
7: Natural Language Processing (NLP)
8: Computer Vision
9: AI Use Cases in IT & Telecom
10: Tools and Platforms
11: Maths in AI
Module 1: Why AI?
Traditional Programming vs. Artificial Intelligence
In traditional programming, you create your program and give your input to that program and program was responsible to generate an output base on the program was designed whereas in Machine learning programming we give input as well as output to the ML algorithm and as a result it will generate a program for you, so that now if you give just input to your program, it will be able to generate output for you.
Machine learning is mainly used to solve complex, real-world problems without explicitly writing the rules or logic — the system figures them out from data.
- What is AI?
Artificial Intelligence (AI) is a branch of computer science that focuses on creating machines that can perform tasks that typically require human intelligence. These tasks include:
✅ Learning – AI learns from data and improves its performance over time. ✅ Reasoning – AI can analyze information and make logical decisions. ✅ Problem-Solving – AI can find solutions to complex problems. ✅ Understanding Language – AI can process and generate human language (like ChatGPT!).
Data Science is the process of cleaning, preparing, and analyzing the data while AI focuses on making machines intelligent using that data.
-
What is Machine Learning?
Machine Learning is all about statistics, its a mathematical calculation.
Machine Learning (ML) is a subset of Artificial Intelligence (AI) that allows computers to learn from data and make decisions or predictions without being explicitly programmed.
Machine Learning is the process by which a machine learns from examples (data) and improves its performance over time. Machine Learning (ML) is about training models on data to make predictions or decisions. It focuses on model development, evaluation, and improvement.
In ML we have three different approaches/models:
-
Supervised ML
-
Unsupervised ML
-
Reinforcement (semi supervised) ML
-
Supervised Learning – In Supervised learning, we will be having a labeled data (past data) and with this kind of data we will be actually will be able to do the prediction for the future.
Supervised Learning – Learning while being guided. The system learns from examples that already have the correct answers (labels), so it knows exactly what it should predict.
Algorithms: • Linear Regression • Logistic Regression • Decision Trees • Support Vector Machines • Neural Networks
- Unsupervised ML –
Unsupervised Learning – Learning by exploring. The system learns from data without any correct answers given, finding patterns and groupings on its own.
Here, we will not be having any labeled data, that means in my data set we will not know what is the output. In this ML model we usually solved clustering kind of problems. What do you mean by clustering? Based on the similarity of the data, it will try to group the data together.
Algorithms: • K-Means Clustering • Hierarchical Clustering • DB scan clustering
- Reinforcement Learning – Some part of data will be labeled and some part of data will not be labeled so the ML model learns slowly by seeing the past data and it will be learning as soon as new data will be coming up.
Reinforcement Learning is a way for a computer or machine to learn by trial and error, getting rewards for good actions and penalties for bad ones, so it can figure out the best way to reach a goal.
"Reinforcement Learning is nothing but learning while doing — the system learns by taking actions, seeing the results, and adjusting its behavior to get better rewards over time."
Algorithms: • Q-learning • Deep Q-Networks (DQN) • Policy Gradient methods
Types of Machine Learning Examples
1)Supervised Learning Examples: Spam detection (Email is spam or not) House price prediction Image classification (Cats vs. Dogs)
2)Unsupervised Learning Examples: Customer segmentation in marketing Anomaly detection (fraud detection in banking)
Reinforcement Learning Examples: Self-driving cars Game-playing AI (e.g., AlphaGo)Robotics
What is Deep Learning?
Biological Neuron:
Artificial Neuron:
Biological Neuron (1st Diagram) • Dendrites → Work like input wires. They receive signals from other neurons. • Cell Nucleus (Soma) → The decision center. It processes incoming signals. • Axon → Like a transmission cable. It carries the signal forward. • Synapse → The connection point where signals are passed to the next neuron. 👉 In short: Biological neuron = receives signals → processes them → sends output.
Artificial Neuron (2nd Diagram) • Inputs (X1, X2, X3, X4) → Just like dendrites, they bring information into the neuron. • Weights (W1, W2, …) → Each input has importance (weight). Bigger weight = stronger influence. • Summation + Bias → Like the nucleus, it combines all inputs using the formula: z=∑(wi⋅xi)+b • Activation Function (f(x)) → Like the neuron’s firing decision: should it activate (send signal) or not? • Output (ŷ) → Final signal passed to the next layer (like an axon carrying signal forward). 👉 In short: Artificial neuron = takes inputs → multiplies with weights → adds bias → applies activation → produces output.
✅ Biological vs Artificial Neuron (Analogy) • Dendrites = Inputs (X) • Synapse strength = Weights (W) • Nucleus (processing) = Summation + Bias • Neuron firing = Activation function • Axon = Output
Deep Learning (in Simple Words) • Deep Learning is a special part of Machine Learning. • Instead of you manually deciding what features (like edges in an image, keywords in text) are important → the neural network automatically learns those features by itself. • It does this using layers of artificial neurons stacked on top of each other. • Each layer learns different levels of features (hierarchies): ◦ First layers → simple things (edges, colors, shapes). ◦ Middle layers → combinations (eyes, nose, wheels). ◦ Final layers → full objects (cat, car, human face).
✅ In One Line Deep Learning = Machine Learning using neural networks, where the system automatically learns simple-to-complex features step by step, instead of us manually designing them.
We have Machine Learning Algorithms, then why we need Deep Learning?
In Machine Learning Algorithms we actually do statistical analysis on dataset whereas in Deep Learning, it uses "Neural networks" to analyze data not any algorithms.
(ML)Algorithms uses set of formula i.e. mathematical formula to analyze data whereas in "Neural networks" we have "Neurons", where we create different different connections i.e. Neural networks, and create a "Multi-layer Neural Network" and this network uses to analyze data.
NLP
NLP is a subfield of AI that focuses on how computers can understand, interpret, and generate human language (text or speech). It uses techniques from both Machine Learning (ML) and Deep Learning (DL).
Natural Language Processing (NLP) provides a way for machines to understand and process human language, and it bridges the gap between human communication and machine learning models.
"NLP acts as an interface layer between human language and ML/DL models — enabling machines to understand, interpret, and respond to natural language."
Think of NLP as a language expert and ML/DL as the brains that need structured data. NLP prepares and translates messy human language into a format ML/DL can process.
🔹 What is NLP (Natural Language Processing)?
NLP is the branch of AI that deals with human language (text or speech). Example: Google Translate, ChatGPT, Alexa. It helps computers read, understand, and generate language like humans. 👉 In simple words: NLP is how computers "learn our language" (words, sentences, grammar, meaning).
🔹 What is Computer Vision (CV)?
CV is the branch of AI that deals with images and videos. Example: Face recognition, self-driving car cameras, medical image analysis. It helps computers see, recognize, and interpret visual information like humans do with eyes. 👉 In simple words: CV is how computers "learn to see" (shapes, objects, movements).
🔹 How NLP Transforms Input into Model-Understandable Language
Raw Input: Human text (e.g., "The cat is sleeping"). Preprocessing: Break text into tokens (words/subwords). "The", "cat", "is", "sleeping" → [tokens] Vectorization (Embedding): Convert tokens into numbers (vectors). Example: "cat" → [0.45, -0.12, 0.88, ...] (a numerical list) Model Input: The model understands only numbers, so it processes these vectors to find meaning.
👉 Words become mathematical vectors that carry meaning.
🔹 How CV Transforms Input into Model-Understandable Language
Raw Input: An image (e.g., a cat photo). Pixel Conversion: Each image is made of pixels (numbers for color & brightness). Example: 256×256 image → a table of numbers. Feature Extraction: The model detects patterns (edges, shapes, colors). e.g., round shape → "eye", triangle → "ear". Model Input: These features are represented as vectors (numbers).
👉 Images become mathematical tensors that represent objects and patterns.
🔹 Common Bridge
Both NLP and CV convert raw human input (words or images) into numbers (vectors/matrices/tensors). AI models only understand numbers and patterns, not words or pictures directly.
✨ In one line: NLP = teaching computers human language by turning words into numbers. CV = teaching computers vision by turning images into numbers.
What is Artificial Intelligence?
Artificial intelligence (AI) refers to the field of computer science focused on developing intelligent machines that can perform complex tasks, such as analyzing, reasoning, and learning that would typically require human intelligence. AI systems are designed to perceive their environment, reason and learn from data, and make decisions or take actions to achieve specific goals.
Explainable AI (XAI) refers to a set of techniques and processes that help you understand the rationale behind the output of a machine learning algorithm. With XAI, you can meet regulatory requirements, improve and debug your models, and have more trust in your AI models’ decisions and predictions.
As stated above, AI includes machine learning. It also encompasses other AI and data science techniques such as deep learning, natural language processing, computer vision, and robotics as described below.
Key AI techniques:
Machine Learning: Using algorithms to enable computers to learn from and make predictions or decisions based on data. As machine learning algorithms are exposed to more data through training, their performance enhances over time. The learning acquired from applying these algorithms on the training data culminates in the formation of machine learning models.
Generative AI: A branch of artificial intelligence that focuses on creating new content, such as images, music, or text, that resembles the style and substance of its training data, but is entirely unique.
Natural Language Processing: Enabling computers to understand, interpret, and generate human language.
Computer Vision: Teaching computers to understand and analyze visual data, such as images or videos.
Deep Learning: Employing neural networks that function with multiple layers like a human brain to learn complex patterns and representations from data.
Reinforcement Learning: Training agents to make sequential decisions by receiving feedback from their environment.
Knowledge Representation and Reasoning: Representing information in a structured manner and using logical rules for problem-solving and decision-making.
Expert Systems: Building computer programs that emulate human expertise in a specific domain to provide intelligent recommendations or solutions.
Robotics: Combining AI with robotics to develop intelligent machines capable of perceiving, interacting, and manipulating the physical world.
Genetic Algorithms: Applying evolutionary principles to solve complex problems by iteratively generating and refining solutions.
Fuzzy Logic: Handling uncertainty and imprecise information by using degrees of truth rather than strict binary values.