Machine Learning

By Dr Agha Ali Raza
Coming Soon

Duration

4 months

Language

Bilingual

Course Access

Eternity

Certificate

On Completion

Machine Learning

By Dr Agha Ali Raza
Coming Soon

About this Course

In Machine Learning (ML), computers efficiently adapt data and solve new problems that are related to previously encountered problems. This allows machines to perform useful exploratory and predictive tasks such as natural language understanding and anomaly detection, without being explicitly programmed. This course covers the theoretical foundations and practical applications of ML and the design, implementation, and analysis of various ML algorithms.

What will you learn

You will learn how to choose, design, and implement the most appropriate ML algorithms for various problem types. If you know a programming language, probability, and some linear algebra, this course is for you! It can adequately prepare you for a career in both industry & academia.

Partners

You may have heard the phrase, a man is known by the company it keeps, similarly a data set is known by the company it keeps!

Course Outline

Week 1
Course Overview
You will be introduced to the traditions and history of Machine Learning as well as a comparison of Artificial intelligence and Machine Learning. In addition, you will also learn about the different challenges and opportunities present in Machine Learning.
Week 2
Supervised Learning
You will learn about features, labels, training, testing etc. and formalizing the supervised learning setup. This includes feature spaces, vectors and label spaces as well as hypothesis spaces and how to evaluate them.
Week 3
KNN Model
You will explore the K-Nearest Neighbour (KNN) classifier model – from the basics to KNN distance and similarity measures and constraints, to KNN errors bounds, enhancements and the curse of dimensionality. This week will provide a holistic approach to this KNN model.
Week 4
Evaluation Matrix
In this week you will learn about the evaluation matrices such as performance and confusion as well as the need for a combined measure. This will be enhanced further by developing an understanding of multiclass classification, gold labels and annotation of data and evaluation of classifiers through comparison and imbalances.
Week 5
Linear Regression
This week’s lesson covers linear regression and delves deeper into the motivation for linearity, the fundamentals of regression, the gradient descent algorithm, the multivariate linear regression and its practical issues.
Week 6
Logistic Regression
You will discover the relation between features/cues/events and some outcome and learn how to implement and train a Logistic Regression algorithm which is the baseline supervised machine learning algorithm for classification.
Week 7
Neural Networks
Building on the concepts of the previous week, we will use Logistic Regression to develop a Neuron and eventually implement and train a Neural Network.
Week 8
Bayes Theorem
This week will review joint and conditional probability, and use these concepts for the derivation of the Bayes Theorem. These will be used to find the Maximum a posteriori (MAP) and Maximum Likelihood Estimation (MLE) for a given dataset.
Week 9
Support Vector Machines
You will explore the different kinds of Support Vector Machines and their applications for the classification and regression analysis of data.
Week 10
Decision Trees
Extending the conclusions from Bayes Theorem and Support Vector Machines, you will now learn to implement Decisions Trees which are used to make predictions about a given dataset.
Week 11
Bagging and Boosting
This week, we will be discussing and analysing Ensemble Methods such as Bagging and Boosting which are used to further refine the predictions of our algorithms.
Week 12
Ensemble Methods
In the last few weeks of the course, we will continue our discussion and analysis of Ensemble Methods such as Random Forests, Gradient Boosted Trees and AdaBoost. These are used to build the “intuition” of our algorithms.
Week 13
Unsupervised Learning
We will now learn the techniques to setup unsupervised learning for our ML algorithms as well as overview use cases of unsupervised learning such as clustering and anomaly detection.
Week 14
Ethics, Fairness, Explainability
We conclude the course with a discussion on the Ethics and Fairness of using AI and ML algorithms as well as recent issues and opportunities which have cropped up as a result of advancements in AI.

Instructors

Dr. Agha Ali Raza

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

You might also like

Course

سیکھنا سیکھیں : کچھ بھی نیا کیسے سیکھ سکتے ہیں   

ڈاکٹر صبیح انور، ڈاکٹر ہانیہ اعظم، ڈاکٹر ارشد احمد اور ڈاکٹر سلیمان شاہد
  • PKR 2500
  • 4 ہفتے
  • 01/05/2023
Course
TEACHING PRACTICE: FUNDAMENTALS OF THE 5E MODEL

Teaching Practice: Fundamentals of the 5E Model

Dr Shazia Awan, Sabahat Quadri, Salahuddin Azmat Isa
  • PKR 4999
  • 4 - 6 weeks
  • 01/05/2023
Course
Course Design & Development

Course Design & Development

Dr Launa Gauthier & Talha Dogar
  • PKR 6751
  • Coming Soon
Course

Health & Wellness – A Manual

Dr Basmaa Ali
  • PKR 5250
  • Coming Soon
Previous
Next

Related Masterclasses

Explore all >
< Back to LUMS

About Us

Schools

Centers

2023 (C) Lahore University of Management Sciences. All Rights Reserved
Facebook-f Instagram Youtube Twitter Linkedin-in