MACHINE

LEARNING
Dr. Agha Ali Raza

Machine Learning Syllabus 01

LUMSx is the center for online learning and
professional development at LUMS. We extend
LUMS’ excellence in teaching and research
beyond its borders by leveraging technology and
innovative pedagogy. Our courses aim to bridge
critical knowledge & skill gaps for Pakistani
learners and to meet their diverse learning
needs, we offer Massive Open Online Courses
(MOOCs), Hybrid Courses, Synchronous (Live)
Courses, and Free Open Online Courses
(OpenCourseWare). We intend to harness
technology for enhancing access, improving
educational quality, and amplifying education’s
impact.

Machine Learning Syllabus 02

Course Format: Self-Paced
Language: English, Urdu

Starting Date: Available Now

Duration: 3 months

Price: PKR. 15,000

Early Bird Discount: PKR. 11,250

Machine Learning Syllabus 03

ABOUT THIS COURSE

Are you intrigued by the world of Artificial Intelligence (AI) and wish to learn about the
fascinating field of Machine Learning (ML)? With applications ranging from self-driving cars to
voice assistants and large language models, Machine Learning is revolutionizing the way we
interact with the world at a fast-evolving pace.

This course is designed to equip you with the essential skills, concepts, and applications of
machine learning, setting you on the path to becoming proficient in this field. In this structured
and immersive course, you’ll go through fundamental concepts to advanced techniques,
guided by a logical progression through eleven well-curated modules. Whether you are diving
into the nuances of supervised learning, grasping the principles behind neural networks, or
exploring the ethical dilemmas encompassing AI and ML, this course provides a
comprehensive learning experience.

With captivating videos, hands-on exercises, and peer and staff feedback, you will be able to
apply machine learning concepts to real-world scenarios. By the end of this course, you’ll not
only have a deep understanding of machine learning techniques but also know how to leverage
them responsibly and ethically in various fields.

This course is led by Dr. Agha Ali Raza, known for his stimulating teaching style and ability to
deconstruct some of the most complex ML algorithms into everyday, applicable concepts.
Let’s embark on this enriching learning journey together, paving your way to becoming a
proficient machine learning practitioner!

Machine Learning Syllabus 04

WHAT WILL YOU LEARN

Course Objective: Ignite enthusiasm for Machine Learning and equip learners with the
foundational skills to harness its potential.

By the end of this machine learning course, learners will be able to:
Intuitively grasp the core principles behind Machine Learning models, tools, and
methodologies.
Master the mathematical underpinnings of statistical learning.
Rigorously navigate the lifecycle of designing, executing, and assessing key Machine
Learning models.
Select the optimal algorithm for specific challenges and discern the merits and limitations
of each.
Comprehend the holistic integration of ML in application areas, spanning data sourcing,
annotation, algorithm selection, societal biases, model explainability, and its
transformative implications.

Machine Learning Syllabus 05

MEET YOUR INSTRUCTOR

Course Instructor

DR. AGHA ALI RAZA

Assistant Professor,
Syed Babar Ali School of
Science & Engineering,
LUMS

Dr. Agha Ali Raza is an Assistant Professor in the Department of Computer Science at LUMS and the founding
director of the Center for Speech and Language Technologies (CSaLT). He is a Fulbright Scholar and received his
Ph.D. from the Language Technologies Institute, School of Computer Science at Carnegie Mellon University,
Pittsburgh, USA.

His research interests include Speech & Natural Language Processing, Speech-based Human Computer
Interfaces, and Information & Communication Technologies for Development (ICT4D). The aim of his research is
to enable information access and social connectivity for under-connected and under-served populations (low-
literate, low-income, tech naïve, visually impaired, linguistically/socially marginalized, and geographically remote
communities, and oral cultures) throughout the developing world. His work in Speech and Language technologies
is focused on the localization of linguistic resources and techniques to Pakistani context for the development of
Speech Recognition, Text-to-Speech, Voice-biometrics, Spoken Term Detection and relevant capabilities. I teach
Natural Language Processing, Speech Processing and Machine Learning at graduate and undergraduate levels.

Dr. Raza’s research has been funded by prestigious organizations like Google Inc., Facebook Research, UNICEF,
GIZ, National Institutes of Health (NIH), the National Academies of Sciences, Engineering, Medicine, Keck Futures
Initiative (NAKFI), and the Higher Education Commission of Pakistan. He is also an Associate Chair in the
program committee for CHI.

Machine Learning Syllabus 06

COURSE OUTLINE

MODULE LEARNING OUTCOMES

Welcome to the course on Machine Learning! In this module you will learn about
what Machine Learning is? Who is this course for? What this course contains and
Welcome to how will you be able to benefit from this course. This introductory module will give
Machine Learning you information on the instructor’s profile, course syllabus and objectives, different
features of the course, grading policies, expectations around academic honesty,
frequently asked questions, and a chance to chat with your peers.

This module will uncover the wonderful world of machine learning, demonstrating
its ubiquity in our lives and explaining its underlying concepts. Through a mix of
theory and examples, this module will give you a comprehensive understanding of
Introduction to
machine learning’s key concepts, historical background, applications, challenges
Machine Learning and how it can be harnessed for social good. The module will also give you an
opportunity to learn the basics of python and apply them through a programming
assessment.

Supervised learning is one of the fundamental techniques in Machine Learning.

This module will equip you with the foundational knowledge and practical skills
Supervised necessary to apply supervised learning algorithms to real-world problems. Through
Learning a combination of theoretical concepts and hands-on exercises, you will gain a solid
understanding of the principles, algorithms, and evaluation methods involved in
supervised learning.

K-NN is a non-parametric method used for both classification and regression

tasks. This module will familiarize you with the underlying principles,
implementation, and evaluation of the K-NN algorithm. Through theoretical
KNN explanations and practical examples, you will gain proficiency in applying K-NN to
real-world problems, selecting an appropriate value for K, handling distance
metrics, dealing with imbalanced data, and optimizing model performance.

Machine Learning Syllabus 07

MODULE LEARNING OUTCOMES

The module provides a comprehensive understanding of essential evaluation

metrics for classification tasks. You will begin with learning about accuracy, build
up to precision, recall, and F1-score, which are widely used performance measures
that assess the effectiveness of classifiers in predicting class labels. This module
will equip you with the knowledge and skills to calculate and interpret these
Evaluation of
metrics accurately. You will gain a solid understanding of the concepts behind
Classifiers precision (the proportion of correctly predicted positive instances), recall (the
proportion of actual positive instances correctly predicted), and F1-score (a
harmonic mean of precision and recall). Through practical examples and exercises,
you will learn how to apply these metrics to assess classifier performance and
make informed decisions based on their results.

In this module, you will gain a comprehensive understanding of linear regression, a

widely-used technique in predictive modeling. You will learn the fundamental
principles and assumptions of linear regression, including linearity and
independence. The module will also focus on parameter estimation, coefficient
Linear
interpretation, and prediction. Additionally, important topics like regularization
Regression techniques will be explored. Through hands-on exercises and real-world datasets,
you will develop practical skills in building, evaluating, and improving linear
regression models, enabling you to analyze data, make accurate predictions, and
extract valuable insights.

Logistic regression is a powerful tool used to predict the probability of a binary

outcome based on a set of input variables. In this module you will cover the
underlying concepts and assumptions of logistic regression, including the logistic
function and loss function. You will also explore the process of model fitting,
Logistic
parameter estimation, and interpretation of results. Practical examples and hands-
Regression on exercises are included to enhance your understanding and application of
logistic regression in real-world scenarios. By the end of the module, you will have
a solid foundation in logistic regression and you will be equipped to utilize this
technique for predictive modeling and decision-making tasks.

The module on Neural Networks provides you with an introduction to this powerful
machine learning technique that mimics the structure and functioning of the
human brain. Neural networks are composed of interconnected nodes, or artificial
Neural
neurons, organized in layers that process and transform data. Here you will cover
Networks the fundamental concepts and components of neural networks, including
activation functions, weight initialization, forward and backward propagation, and
gradient descent optimization.

Machine Learning Syllabus 08

MODULE LEARNING OUTCOMES

The module on Support Vector Machines (SVM) offers an introduction to this

powerful supervised learning algorithm used for classification and regression
tasks. SVMs aim to find the optimal hyperplane that separates data points of
Support Vector
different classes with the largest margin. In this module you will learn about the
Machines underlying principles of SVM, including the concept of support vectors, kernel
functions, and the margin optimization objective. You will explore both linear and
nonlinear SVMs, highlighting their strengths and limitations.

This module provides you with an introduction to Bayes Theorem, a fundamental

concept in probability theory and statistics. Bayes Theorem allows us to update our
beliefs about the probability of an event based on new evidence or information.
Bayes
The content sheds light on the core components of Bayes Theorem, including prior
Theorem probabilities, likelihoods, and posterior probabilities. It also explores how Bayes
Theorem can be applied to various scenarios, such as medical diagnostics and
spam filtering.

This module introduces the Naive Bayes classifier, a simple yet effective
probabilistic algorithm used for classification tasks. The Naive Bayes classifier is
Naive Bayes based on Bayes’ theorem and makes the assumption of independence among
Classifier features. Here, you will cover the key concepts and workings of the Naive Bayes
classifier, including the calculation of prior probabilities, likelihoods, and posterior
probabilities.

This module aims to unveil the ‘black box’ nature of artificial intelligence and
machine learning models, enabling deeper understanding of their inner workings
Responsible AI and addressing the multifaceted issues related to AI ethics, fairness and
and Machine explainability. It covers fairness in AI, interpretability of ML models, sources of bias
Learning for and techniques to mitigate bias. The module also touches upon ethics in AI to
understand the moral principles guiding AI development and its use. Lastly, the
Development content covers machine learning for development, explaining how ML techniques
can be used to address social and economic challenges in developing countries.

Machine Learning Syllabus 09

MACHINE
LEARNING
ENROLL NOW

As an online learning gateway of LUMS, we aim to

extend LUMS’ excellence in teaching and research
beyond its borders by leveraging technology.

Machine Learning Syllabus 10