Introduction to AI

Machine Learning
Course Outline and Overview
Unit I: Introduction to AI

1. What is AI
2. Components of AI
3. Introduction to ML
4. Types of ML
5. Influence of AI-ML on Business
6. How to Get Started on AI
7. A Typical AI & ML Process
8. Introduction to Deep Learning
What is AI?

AI is the simulation of human intelligence in machines that are programmed to

think and learn.

Examples: Speech recognition, problem-solving, learning, and planning.

 1. Learning
● Definition: The ability of an AI system to acquire and improve knowledge or skills from experience or
data.

● Types:

○ Supervised Learning

○ Unsupervised Learning

○ Reinforcement Learning

● Example: A spam filter that gets better over time by analyzing more emails.
 2. Reasoning
Definition: The ability to make logical deductions and draw conclusions from given information.

Types:

● Deductive reasoning (from general rules to specific conclusions)

● Inductive reasoning (from specific cases to general rules)

Example: AI diagnosing a disease based on symptoms and medical knowledge.

 3. Problem Solving
Definition: AI’s ability to identify solutions to complex tasks or challenges by evaluating different options.

Techniques:

● Search algorithms

● Optimization

● Heuristics

Example: Chess-playing AI determining the best move among many.

 4. Perception
Definition: The process of acquiring and interpreting sensory information from the environment.

Areas:

● Computer Vision (seeing)

● Speech Recognition (hearing)

● Touch Sensors (tactile input)

Example: Self-driving cars identifying pedestrians and traffic signs using cameras.
 5. Language Understanding
Definition: AI's ability to understand, interpret, and generate human language.

Subfields:

● Natural Language Processing (NLP)

● Machine Translation

● Sentiment Analysis

Example: Virtual assistants like Siri or Alexa understanding voice commands.

Introduction to Machine Learning

Machine Learning is a subset of AI that involves the use of

algorithms and statistical models to perform tasks without
explicit instructions.

Examples: Image recognition, recommendation systems.

Practical Example: Classifying Fruits
Scenario:

You have a basket of fruits, and you want the computer to learn how to identify them.

Color Weight Fruit Type

(grams)

Red 150 Apple

Yellow 120 Banana

Red 180 Apple

Yellow 110 Banana

👉 The model learns from labeled data — it knows the correct answer while training.
📌 After training, when you give it a new fruit (e.g., Red, 160g), it can predict: "Apple".
Example 2: Unlabelled Dataset

Color Weight
(grams) 👉 The model tries to find patterns or groupings on its own.

Red 150
📌 It might group them into two clusters — red and yellow —
Yellow 120
without knowing they are apples and bananas.
Red 180

Yellow 110
Types of Machine Learning

1. Supervised Learning
2. Unsupervised Learning
3. Semi-supervised Learning
4. Reinforcement Learning
1. Supervised Learning

Definition:
Supervised learning is a type of machine learning where the model is trained on
labeled data.

Each training example includes input data and the correct output (label), and the
model learns to map inputs to the correct outputs.

Example:
An email spam filter is trained with emails labeled as “Spam” or “Not Spam.” It
learns to classify new emails based on those examples.
2. Unsupervised Learning

Definition:
Unsupervised learning is a type of machine learning where the model is given
unlabeled data, and it tries to find patterns, groupings, or structures in the data
without knowing the correct answers.

Example:
A shopping app groups users into clusters (e.g., based on spending habits) without
knowing anything about them beforehand. This helps in targeted marketing.
 3. Semi-Supervised Learning

Definition:
Semi-supervised learning is a method that uses a small amount of labeled data and
a large amount of unlabeled data.

It helps improve learning accuracy while reducing the cost of labeling.

Example:
In face recognition, only a few images are labeled with names.

The model uses them to guess names in the rest of the unlabeled face images (used
by social media platforms for auto-tagging).
 4. Reinforcement Learning

Definition:
Reinforcement learning is a type of machine learning where an agent learns to make
decisions by performing actions and receiving rewards or penalties based on those
actions.

It learns through trial and error.

Example:
A self-driving car in a simulation learns to drive by getting rewards for staying on
the road and penalties for crashing. Over time, it learns the best driving strategy.
Influence of AI-ML on Business

Impact:

● Automation
● Data-driven decision-making
● Improved customer experience
● Cost reduction
● New product and service innovation
Machine Learning (ML) is utilized across a wide variety of fields and
industries. Here are some prominent examples:

1. Healthcare

● Disease Diagnosis: ML algorithms can analyze medical images to detect

diseases like cancer.
● Personalized Medicine: Tailoring treatments based on individual genetic
profiles.
● Predictive Analytics: Predicting outbreaks of diseases and patient
readmission rates.
2. Finance

● Fraud Detection: Identifying fraudulent transactions in real-time.

● Algorithmic Trading: Automating trading strategies based on data
analysis.
● Credit Scoring: Assessing the creditworthiness of individuals.
3. Retail

● Customer Recommendations: Recommending products based on past

purchases and browsing history.
● Inventory Management: Predicting stock needs to optimize inventory
levels.
● Personalized Marketing: Tailoring marketing messages to individual
customer preferences.
4. Transportation

● Autonomous Vehicles: Enabling self-driving cars to navigate and make

decisions.
● Route Optimization: Finding the most efficient routes for delivery
services.
● Predictive Maintenance: Predicting when vehicle parts will need
maintenance to prevent breakdowns.
5. Manufacturing

● Quality Control: Detecting defects in products using computer vision.

● Predictive Maintenance: Predicting equipment failures before they
happen.
● Supply Chain Optimization: Improving logistics and supply chain
efficiency.
6. Entertainment

● Content Recommendation: Suggesting movies, music, and shows based

on user preferences.
● Content Creation: Using ML to create music, art, and even news
articles.
● Audience Analysis: Understanding viewer preferences and behavior.
7. Agriculture

● Crop Monitoring: Using drones and satellite images to monitor crop

health.
● Yield Prediction: Predicting crop yields based on various factors like
weather and soil conditions.
● Precision Farming: Optimizing the use of resources like water and
fertilizers.
8. Energy

● Demand Forecasting: Predicting energy consumption patterns to

optimize production.
● Renewable Energy Management: Managing the integration of
renewable energy sources into the grid.
● Predictive Maintenance: Predicting when energy infrastructure needs
maintenance.
9. Telecommunications

● Network Optimization: Optimizing network performance and

reliability.
● Customer Service: Using chatbots to handle customer inquiries.
● Churn Prediction: Identifying customers likely to leave and taking
proactive measures.

10. Human Resources

● Recruitment: Screening resumes and identifying the best candidates.

● Employee Retention: Predicting which employees are likely to leave.
● Performance Management: Analyzing employee performance and
providing feedback.
11. Education

● Personalized Learning: Adapting educational content to individual

learning styles.
● Administrative Tasks: Automating administrative tasks like grading and
scheduling.
● Predictive Analytics: Identifying students who are at risk of dropping
out.

12. Security

● Surveillance: Analyzing video feeds to detect suspicious activities.

● Threat Detection: Identifying cybersecurity threats in real-time.
● Biometric Authentication: Using facial recognition and other biometrics
for security purposes.
13. Customer Service
● Chatbots: Providing customer support through automated systems.
● Sentiment Analysis: Analyzing customer feedback to improve services.
● Call Center Optimization: Predicting call volumes and optimizing
staffing.
14. Smart Homes
● Voice Assistants: Devices like Amazon Alexa and Google Home that
understand and respond to voice commands.
● Home Automation: Automating home systems like lighting, heating, and
security.
● Energy Management: Optimizing energy usage within the home.
How to Get Started on AI

Learn the basics of programming

Understand key AI concepts

Take online courses and certifications

Work on projects and build a portfolio

Stay updated with the latest trends and research

1. Understand What AI Is

Before diving in, understand the basic concepts of AI:

● Artificial Intelligence (AI): Machines that mimic human intelligence.

● Machine Learning (ML): A subset of AI where machines learn from data.

● Deep Learning (DL): A subset of ML using neural networks.

 2. Learn the Required Prerequisites
📐 Math:
● Linear Algebra (Vectors, Matrices)
● Probability and Statistics
● Calculus (basic understanding)

💻 Programming:
● Python is the most widely used language in AI.
● Learn basic syntax, data structures, and libraries like numpy, pandas, matplotlib.

💾 Data Handling:
● Learn how to read, clean, and manipulate datasets (CSV, Excel, etc.)
 3. Start with Machine Learning
AI is a big field — starting with Machine Learning (ML) is more practical.

📘 Learn ML Topics:

● Supervised Learning (Regression, Classification)

● Unsupervised Learning (Clustering)
● Model Evaluation (accuracy, precision, recall)
● Overfitting, underfitting

🛠 Tools to Use:

● Jupyter Notebook (for Python-based ML)

● scikit-learn (for ML models)
● Google Colab (online coding)
4. Try Mini Projects (Hands-On Practice)

Build small projects to apply what you’ve learned:

● Spam email classifier

● Movie recommendation system

● House price prediction

● Handwritten digit recognition (using MNIST)

💡 Tip: Don’t just copy code — understand each step.

5. Build a Portfolio

● Upload your projects to GitHub

● Participate in hackathons, Kaggle competitions

📂 This helps you build credibility and land jobs or freelance work.
🔧 Tools to Get Started With:

Tool Use Case

Python Programming language

Google Colab Cloud-based notebooks

scikit-learn ML models

TensorFlow Deep Learning

PyTorch Deep Learning

Kaggle Datasets + Competitions

GitHub Project hosting + portfolio

 Predicting Student Marks Using Machine Learning

● This beginner-friendly code demonstrates how machine learning can be used to predict
outcomes based on past data.
● Using a simple Linear Regression model, it predicts student marks based on the number
of study hours.
● The model is trained on a small dataset and makes predictions while also visualizing the
relationship between input and output with a graph.

https://colab.research.google.com/drive/1B89BJz0jeu8G8os6CEMQTbPcfIJq3-AC?usp=sharing
A Typical AI & ML Process
(Machine Learning Lifecycle)

How to choose and build the right machine learning model?

1. Problem definition
2. Data collection and preprocessing
3. Model selection and training
4. Model evaluation
5. Deployment
6. Monitoring and maintenance

PDM²DM
 Problem Identification and Objective Setting

By framing the problem in a comprehensive manner, the team establishes a foundation

for the entire machine learning lifecycle. Crucial elements, such as project objectives,
desired outcomes, and the scope of the task, are carefully delineated during this stage.
Here are the basic features of problem definition:
● Collaboration: Work together with stakeholders to understand and define the
business problem.
● Clarity: Clearly articulate the objectives, desired outcomes, and scope of the
task.
● Foundation: Establish a solid foundation for the machine learning process by
framing the problem comprehensively.
 Data Collection
This phase involves the systematic gathering of datasets that will serve as the raw material
for model development. The quality and diversity of the data collected directly impact the
robustness and generalizability of the machine learning model.
Here are the basic features of Data Collection:
● Relevance: Collect data that is relevant to the defined problem and includes
necessary features.
● Quality: Ensure data quality by considering factors like accuracy, completeness,
and ethical considerations.
● Quantity: Gather sufficient data volume to train a robust machine learning model.
● Diversity: Include diverse datasets to capture a broad range of scenarios and
patterns.
 Data Preprocessing
Preprocessing takes this a step further by standardizing formats, scaling values, and encoding categorical
variables, creating a consistent and well-organized dataset. The objective is to refine the raw data into a
format that facilitates meaningful analysis during subsequent phases of the machine learning lifecycle. By
investing time and effort in data cleaning and preprocessing, practitioners lay the foundation for robust
model development, ensuring that the model is trained on high-quality, reliable data.
Here are the basic features of Data Cleaning and Preprocessing:
● Data Cleaning: Address issues such as missing values, outliers, and inconsistencies in the data.
● Data Preprocessing: Standardize formats, scale values, and encode categorical variables for
consistency.
● Data Quality: Ensure that the data is well-organized and prepared for meaningful analysis.
● Data Integrity: Maintain the integrity of the dataset by cleaning and preprocessing it
effectively.
 Model Selection
Model selection is a pivotal decision that determines the algorithmic framework guiding the predictive
capabilities of the machine learning solution. The choice depends on the nature of the data, the complexity
of the problem, and the desired outcomes.
Here are the basic features of Model Selection:
● Alignment: Select a model that aligns with the defined problem and characteristics of the dataset.
● Complexity: Consider the complexity of the problem and the nature of the data when choosing a
model.
● Decision Factors: Evaluate factors like performance, interpretability, and scalability when
selecting a model.
● Experimentation: Experiment with different models to find the best fit for the problem at hand.
 Model Training
Model training is an iterative and dynamic journey, where the algorithm adjusts its parameters to minimize
errors and enhance predictive accuracy. During this phase, the model fine-tunes its understanding of the
data, optimizing its ability to make meaningful predictions. Rigorous validation processes ensure that the
trained model generalizes well to new, unseen data, establishing a foundation for reliable predictions in
real-world scenarios.
Here are the basic features of Model Training:
● Training Data: Expose the model to historical data to learn patterns, relationships, and
dependencies.
● Iterative Process: Train the model iteratively, adjusting parameters to minimize errors and
enhance accuracy.
● Optimization: Fine-tune the model’s understanding of the data to optimize predictive capabilities.
● Validation: Rigorously validate the trained model to ensure generalization to new, unseen data.
 Model Evaluation
Evaluation is a critical checkpoint, providing insights into the model’s strengths and weaknesses. If the model falls
short of desired performance levels, practitioners initiate model tuning—a process that involves adjusting
hyperparameters to enhance predictive accuracy. This iterative cycle of evaluation and tuning is crucial for achieving
the desired level of model robustness and reliability.
Here are the basic features of Model Evaluation and Tuning:
● Evaluation Metrics: Use metrics like accuracy, precision, recall, and F1 score to evaluate model
performance.
● Strengths and Weaknesses: Identify the strengths and weaknesses of the model through rigorous testing.
● Iterative Improvement: Initiate model tuning to adjust hyperparameters and enhance predictive accuracy.
● Model Robustness: Iterate through evaluation and tuning cycles to achieve desired levels of model
robustness and reliability.
 Deployment and Integration

Model deployment marks the culmination of the machine learning lifecycle, transforming theoretical
insights into practical solutions that drive tangible value for organizations.
Here are the basic features of Model Deployment:
● Integration: Integrate the trained model into existing systems or processes for real-world
application.
● Decision Making: Use the model’s predictions to inform decision-making and drive tangible
value for organizations.
● Practical Solutions: Deploy the model to transform theoretical insights into practical solutions
that address business needs.
● Continuous Improvement: Monitor model performance and make adjustments as necessary to
maintain effectiveness over time.
 Monitoring and Maintenance

After deploying the model to production we need to constantly monitor and improve the
system. We will be monitoring model metrics, hardware and software performance, and
customer satisfaction.

The monitoring is done completely automatically, and the professionals are notified
about the anomalies, reduced model and system performance, and bad customer
reviews.

After we get a reduced performance alert, we will assess the issues and try to train the
model on new data or make changes to model architectures. It is a continuous process.
Introduction to Deep Learning

Definition: A subset of ML that uses neural networks with three or more

layers.

Examples: Image classification, natural language processing.

 Deep Learning
● The deep learning is a model based on Artificial Neural Networks (ANN), more specifically
Convolutional Neural Networks (CNN)s.
● There are several architectures used in deep learning such as deep neural networks, deep
belief networks, recurrent neural networks, and convolutional neural networks.
● These networks have been successfully applied in solving the problems of computer vision,
speech recognition, natural language processing, bioinformatics, drug design, medical image
analysis, and games.
● There are several other fields in which deep learning is proactively applied.
● The deep learning requires huge processing power and humongous data, which is generally
easily available these days.
 What is a Neural Network?

● Neural Networks are inspired by the most complex object in the universe – the
human brain. Let us understand how the brain works first. The human brain is
made up of something called Neurons. A neuron is the most basic
computational unit of any neural network, including the brain.
● Neurons take input, process it, and pass it on to other neurons present in the
multiple hidden layers of the network, till the processed output reaches the
Output Layer.
● In its simplest form, an Artificial Neural Network (ANN) has only three layers –
the input layer, the output layer, and a hidden layer.