Principles of Programming Using C

Learning Center Of Excellence With Google

Department of Computer Applications

Lecture - 10

Course Coordinator:-
Name- Ms. Urvashi Sharma Course Name:- BCA I Sem.
Course Code:- CA 1004
Designation- Assistant Professor
Mail Id : [email protected]
 Key Points

S.No. Highlights

1 Objective of the Topic

2 Introduction

3 Core content of the Topic

Teaching methodology- Case study/ experiential learning/ visual aids/ hands on

4
practice etc.

5 Learning Outcomes of the Topic

6 Summary of the Topic/ Key take away slide

7 Quiz / Practice Questions

8 References
School of Engineering and Technology/ Department of Computer Applications
 Course Objectives

• To introduce the fundamentals of Machine Learning, including its types, real-world applications,
tools, and common challenges in model development and deployment.

• To develop students’ understanding of supervised learning techniques, with a focus on classification

and regression algorithms, and equip them with the ability to evaluate and improve model
performance.

• To familiarize students with unsupervised learning methods, such as clustering and dimensionality
reduction, and demonstrate their practical applications in data exploration and segmentation.

• To provide hands-on experience using ML tools and platforms like Scikit-learn, TensorFlow, and
Google Colab for building, training, and testing machine learning models.

• To cultivate skills in model optimization and deployment, including feature engineering, validation
techniques, and the implementation of end-to-end ML workflows through real-world mini-projects.
 Abstract

• Course Objectives

• Learning Objectives

• Content

• Course Outcomes

• Learning Outcomes

• Questions Answers

• References

• Supporting Study Material for next lecture

 Learning Objectives

• Understand and explain the core concepts of machine learning, including its types, applications, and the
typical steps involved in a machine learning pipeline.

• Differentiate between classification and regression tasks and apply appropriate supervised learning
algorithms such as k-NN, Decision Trees, Logistic Regression, and Linear Regression using Python
libraries.

• Evaluate machine learning models using key performance metrics like accuracy, precision, recall, F1-
score, MSE, RMSE, and R² to assess model effectiveness.

• Implement unsupervised learning techniques such as K-Means, Hierarchical Clustering, and PCA, and
apply them to real-world problems like customer segmentation.

• Apply model optimization techniques including cross-validation, feature scaling, and hyperparameter
tuning, and build an end-to-end machine learning project using real-world datasets.
 Course Outcome

• Explain the fundamentals of machine learning, its types, challenges, and tools used

for model development and deployment.

• Apply supervised learning algorithms for both classification and regression tasks

using appropriate Python libraries.

• Evaluate machine learning models using performance metrics such as accuracy,

precision, recall, F1-score, MSE, RMSE, and R².

• Implement unsupervised learning techniques like clustering and dimensionality

reduction to solve real-world problems.

• Execute complete machine learning workflows including data preprocessing, model

optimization, and deployment using platforms like Google Colab and Scikit-learn.
 Definition
🔹 1. Definition of Machine Learning
• Machine Learning (ML) is a subfield of Artificial Intelligence
(AI) that enables systems to learn from data and make decisions
or predictions without being explicitly programmed.
• Formal Definition (Arthur Samuel, 1959):
"Machine learning is the field of study that gives computers the
ability to learn without being explicitly programmed."
• Key Characteristics:
• Learns from historical data.
• Improves performance over time.
• Reduces human intervention in decision-making.
• Handles large, complex datasets efficiently.
2. How Machine Learning Works (Simplified View)
Input Data – Raw data collected from the environment or databases.
Feature Extraction – Selecting important variables.
Model Selection – Choosing an algorithm (e.g., Linear Regression,
Decision Trees).
Training – Feeding data into the model to learn patterns.
Testing – Validating performance on unseen data.
Prediction – Using the trained model for real-world use.
🔹 3. Types of Machine Learning
Type Description Examples

Supervised Learn from labeled data Spam detection, loan prediction

Unsupervised Learn from unlabeled data Customer segmentation, clustering

Reinforcement Learn via feedback (rewards/punishments) Game playing agents, robotics

Semi-supervised Mix of labeled and unlabeled data Medical imaging, fraud detection
🔹 4. Brief History of Machine Learning

📜 1950s–1970s: Foundations of AI & ML

• 1950: Alan Turing proposes the Turing Test.
• 1957: Perceptron invented by Frank Rosenblatt – first neural network.
• 1967: Nearest Neighbor algorithm introduced.

📜 1980s–1990s: Emergence of ML Algorithms

• Rise of decision trees, support vector machines, Bayesian networks.
• 1995: Random Forest introduced.
• Backpropagation popularized in neural networks.

📜 2000s–2010s: Big Data and Deep Learning

• Massive data and computing power make ML more effective.
• Deep Learning (multi-layered neural networks) achieves state-of-the-art performance.
• Applications in speech recognition, vision, NLP, autonomous driving.
• 📜 2020s: Widespread Adoption
• Integration in healthcare, finance, education, marketing.
• Tools like TensorFlow, Scikit-learn, Google Colab make ML accessible.
• Generative AI (e.g., ChatGPT) takes center stage.
• 🔹 5. Importance of Machine Learning
• 🔸 1. Automates Decision-Making
• Reduces human effort in repetitive or data-intensive tasks.
• Example: Automated fraud detection in banks.
• 🔸 2. Enables Predictive Analytics
• Helps forecast outcomes using historical data.
• Example: Demand forecasting, stock market trends.
• 🔸 3. Enhances Personalization
• Recommender systems (e.g., Netflix, Amazon, YouTube) tailor experiences to users.
• 🔸 4. Real-Time Applications
• ML systems can work in real time: autonomous vehicles, facial recognition, virtual assistants.
• 🔸 5. Improves Efficiency
• Optimizes business operations, supply chains, and customer services.
• 🔸 6. Drives Innovation
• Enables development of intelligent systems across sectors like healthcare (e.g., cancer
detection), agriculture (e.g., crop disease prediction), and education (e.g., adaptive learning).
🔹 6. Real-World Examples
Domain Application

Healthcare Disease prediction, medical image analysis

Finance Fraud detection, credit scoring

Marketing Customer segmentation, churn prediction

Agriculture Crop yield forecasting

Transportation Traffic prediction, self-driving cars

• 🔹 7. Summary Points
• Machine Learning is a powerful subset of AI focused on
learning from data.
• It has evolved from basic pattern recognition to complex
deep learning systems.
• ML is integral to many modern technologies and continues
to revolutionize industries.
Question Answer
 Question Answer
Who is considered the father of Machine Learning for defining it in 1959?
a) Alan Turing
b) Frank Rosenblatt
c) Arthur Samuel
d) Geoffrey Hinton
Answer: → c) Arthur Samuel
 Question Answer

What is the primary goal of Machine Learning?

a) Write programs for every task
b) Automate database design
c) Enable machines to learn from data and improve
performance
d) Replace operating systems
Answer: → c) Enable machines
to learn from data and improve
performance
 Question Answer
Which of the following is a type of Machine Learning?
a) Logical learning
b) Evolutionary learning
c) Supervised learning
d) Symbolic reasoning
→ c) Supervised learning
 Question Answer
Which algorithm is commonly used in supervised learning?
a) K-Means
b) Principal Component Analysis
c) Linear Regression
d) Apriori
Answer: → c) Linear Regression
 Question Answer
Unsupervised learning deals with:
a) Labeled data
b) Structured programs
c) Unlabeled data
d) Reinforcement signals
Answer: → c) Unlabeled data
 Question Answer
Which one of the following is a real-life application of Machine Learning?
a) Social media scrolling
b) Writing HTML code
c) Fraud detection in banks
d) Printing documents
Answer: → c) Fraud detection in
banks
 Question Answer
Which of the following is a key characteristic of Machine Learning?
a) Manual rule definition
b) Direct programming for logic
c) Ability to learn and adapt from data
d) Works only with small datasets
Answer: → c) Ability to learn and adapt from
data
 Question Answer
In which decade was the Perceptron model introduced?
a) 1940s
b) 1950s
c) 1960s
d) 1980s
Answer: → b) 1950s
 Question Answer
Which of the following platforms is widely used for Machine Learning
today?
a) Adobe Photoshop
b) AutoCAD
c) Google Colab
d) MySQL
Answer: → c) Google Colab
 Question Answer
Which technique is typically used in Reinforcement
Learning?
a) Reward and punishment
b) Clustering
c) Feature selection
d) Linear regression
Answer: → a) Reward and
punishment
 Question Answer
The evolution of Deep Learning significantly progressed in which
decade?
a) 1980s
b) 1990s
c) 2000s
d) 2010s
Answer: → d) 2010s
 Question Answer
Which field does Machine Learning NOT typically support?
a) Robotics
b) Agriculture
c) Banking
d) Fossil analysis
Answer: → d) Fossil analysis
 Question Answer
What does the term "training a model" mean in ML?
a) Coding the model using C++
b) Feeding data into an algorithm to learn patterns
c) Storing data in a database
d) Formatting the input values
Answer: → b) Feeding data into
an algorithm to learn patterns
 Question Answer
Which of the following is NOT an ML tool or library?
a) Scikit-learn
b) TensorFlow
c) NumPy
d) Microsoft Excel
Answer: → d) Microsoft Excel
 Question Answer
What is the main reason for ML's popularity in the 21st century?
a) Increase in internet speed
b) Decrease in interest in databases
c) Availability of large datasets and powerful computing
d) Shift from desktop to mobile usage
Answer: → c) Availability of large
datasets and powerful
computing
 Learning Outcomes
• Define Machine Learning and distinguish it from traditional programming and general AI.

• Explain the three main types of Machine Learning: Supervised, Unsupervised, and Reinforcement
Learning.

• Summarize the historical evolution of Machine Learning, highlighting key milestones and
contributors.

• Identify real-world applications and the growing importance of Machine Learning across industries.

• Describe the basic steps involved in a Machine Learning pipeline from data collection to prediction.
 References
Text Books:
• Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow
Author: Aurélien Géron
Publisher: O’Reilly Media
Edition: 2nd Edition, 2019
ISBN: 9781492032649
• Pattern Recognition and Machine Learning
Author: Christopher M. Bishop
Publisher: Springer
Edition: 1st Edition, 2006
ISBN: 9780387310732
• Introduction to Machine Learning with Python: A Guide for Data Scientists
Authors: Andreas C. Müller and Sarah Guido
Publisher: O’Reilly Media
Edition: 1st Edition, 2016
ISBN: 9781449369415
• Machine Learning: A Probabilistic Perspective
Author: Kevin P. Murphy
Publisher: MIT Press
Edition: 1st Edition, 2012
ISBN: 9780262018029
 References
Reference Books:
• Machine Learning
Author: Tom M. Mitchell
Publisher: McGraw-Hill Education
Edition: 1st Edition, 1997
ISBN: 9780070428072
Description: A foundational text that provides theoretical insights and basic algorithms in machine learning.
• Understanding Machine Learning: From Theory to Algorithms
Authors: Shai Shalev-Shwartz and Shai Ben-David
Publisher: Cambridge University Press
Edition: 2014
ISBN: 9781107057135
Description: Focuses on the mathematical foundations of machine learning and theoretical guarantees of algorithms.
• Machine Learning Yearning
Author: Andrew Ng
Publisher: Self-published (available online)
Description: A practical guide focused on structuring ML projects, written by one of the pioneers in AI.
• Data Mining: Concepts and Techniques
Authors: Jiawei Han, Micheline Kamber, Jian Pei
Publisher: Morgan Kaufmann
Edition: 3rd Edition, 2011
ISBN: 9780123814791
Description: Covers data mining techniques and concepts relevant to machine learning applications.
• Deep Learning
Authors: Ian Goodfellow, Yoshua Bengio, and Aaron Courville
Publisher: MIT Press
Edition: 1st Edition, 2016
ISBN: 9780262035613
Description: Provides an in-depth understanding of deep learning, useful for advanced study beyond classical ML.
 References
🎓 NPTEL Course (Text-Based)
“Introduction to Machine Learning” by IIT Madras
Available on the NPTEL/SWAYAM portal, this course covers core ML paradigms, including
supervised/unsupervised learning and reinforcement learning
youtube.com+15nptel.ac.in+15reddit.com+15
onlinecourses.nptel.ac.in+3onlinecourses.nptel.ac.in+3en.wikipedia.org+3
📌 Course Link: https://nptel.ac.in/courses/106106139 (accessible via SWAYAM)

• 🎥 YouTube Reference – Lecture Series

Week 1 – Lecture 1: Introduction to Machine Learning (by the same IIT Madras course)
 Supporting Study Material for the next
lecture

🔢 Syllabus Units:
• Introduction to Machine Learning
(Overview, types, applications, pipeline, tools like Scikit-learn, TensorFlow)

• Supervised Learning – Classification

(k-NN, Decision Trees, Naïve Bayes, Logistic Regression, Confusion Matrix, etc.)

• Supervised Learning – Regression

(Linear Regression, Lasso, Ridge, MSE, R², etc.)

• Unsupervised Learning
(K-Means, Hierarchical, DBSCAN, PCA, customer segmentation)

• Model Optimization and Projects

(Cross-validation, Grid search, Bias-Variance, Feature scaling, Mini-project)
Thank You!