SYLLABUS
COMP 486 is an introductory undergraduate course in machine learning. The class will briefly cover topics in regression, classification, mixture models, neural networks,and ensemble methods .s.
This course uses concrete examples, minimal theory, and production-ready Python frameworks (Scikit-Learn, Keras, and TensorFlow) to help you gain an intuitive understanding of the concepts and tools for building intelligent systems, so we assumes that the students have some Python programming experience. If you don’t know Python yet, https://learnpython.org is a great place to start. The official tutorial on Python.org is also quite good.
Goals
At the conclusion of this course, you will be able to describe the principal models used in machine learning and the types of problems to which they are typically applied, compare the assumption made in each model and the strengths and weakness of each model, determine to which problems machine learning is applicable and which model or models would be most appropriate in each case, apply the principal models in machine learning to appropriate problems. .
About COMP 486
| Prerequisites | COMP 210 (Data Structure). |
|---|---|
| Instructor | Dr. Tasnim Gharaibeh Olds/Upton 203F, (269) 337-7526 Dr. Tasnim (preferred) | Dr. Gharaibeh she, her, hers Office Hours :
|
| Email Rules |
Subject COMP486: Your Subject. Example COMP486: About project |
| Required Text |
|
| Microsoft Team Site | You can download the app and install it in your system here: MS Teams |
Computing Resources and Software
To be able to successfully complete this course, you will be required to have the following tools:
- Programming Language: Python
- Development Environment: We will use the Google Colab
All the code examples in this book are open source and available online at https://git hub.com/ageron/handson-ml3, as Jupyter notebooks. These are interactive documents containing text, images, and executable code snippets (Python in our case). The easiest and quickest way to get started is to run these notebooks using Google Colab: this is a free service that allows you to run any Jupyter notebook directly online, without having to install anything on your machine. All you need is a web browser and a Google account.
Other option You can use Jupyter in Anaconda Navigator, which is a desktop graphical user interface (GUI) included in Anaconda® Distribution that allows you to launch applications and manage conda packages, environments, and channels without using command line interface (CLI) commands. Navigator can search for packages on Anaconda.org or in a local Anaconda Repository. It is available for Windows, macOS, and Linux. You can use the following link to download and install Anaconda Navigator.
Topics to be Covered
The following are the topics we will be learning about and discussing during COMP 486 (and a tentative schedule for when each of them will be covered).
| Week 1 | Introduction to the course
Introduction to ML |
|---|---|
| Week 2 | Machine Learning project Get, Discover and Visualize the Data Select the model |
| Week 3 | Classification Performance Measures |
| Week 4 | Linear Regression Gradient Descent Regularized Linear Models |
| Weeks 5 | Logistic Regression Linear SVM Classification |
| Weeks 6 | Nonlinear SVM Classification Decision Trees |
| Weeks 7 | Ensemble Learning Random Forests Introduction to Artificial Neural Networks |
| Weeks 8 | ANN - The Perceptron The Multilayer Perceptron and Backpropagation Regression MLPs |
| Weeks 9 | Classification MLPs Training Deep Neural Networks |
| Weeks 10 | Project Presentations |
Assignments and Class Evaluation
Assignments, announcements, class notes, and other material will be made available on the course web site: http://www.cs.kzoo.edu/cs486/. Students are responsible for checking this resource frequently.
Reading assignments and discussion questions or exercises may be assigned. You are encouraged to discuss both the ideas from the reading and your solutions to any exercises using MS Teams or email.
There will be approximately 4-5 assignments assigned throughout the quarter, which may take a week or longer to complete. The time required to solve the assignment is difficult to predict, but time-management skills are as critical in industry as they are in college. I will make assignments available online far enough in advance that you will have some flexibility in scheduling your work, but you are responsible for budgeting your time wisely so that you will be able to complete your assignments on time. Assignments must always be turned in on time unless you clear it with me in advance.
There will be one project that could be done in teams of 2-4 students.
There will be one midterm and a final this quarter.
Final Grade
Final grade will be based on:
- Participation: attendance, discussion questions, occasional quizzes, in-class activities 25%
- Assignments: 4 or 5 take home assignments 40%
- Project 25%
- Midterm Exam 5%
- Final Exam 5%
Collaboration and the Honor System
This course operates in accordance with the principles of the Kalamazoo College Honor System: responsibility for personal behavior, independent thought, respect for others, and environmental responsibility. In particular, academic integrity is a fundamental principle of scholarship. Representing someone else's work as your own, in any form, constitutes academic dishonesty. Unauthorized collaboration and receiving help from others outside the bounds permitted by the instructor are also violations of the College honor system. You are responsible for working within the permitted bounds, and acknowledging any help from others or contributions from other sources.
Exams and quizzes should be entirely your own work, unless specified.
Late submission policy: Assignment due dates have two important functions: to help you plan your time and keep you on track to successfully complete the course, and to make grading more manageable. Late assignments will accrue late penalties or might not be accepted at all. To encourage timeliness, assignments that are one day late will lose 2%; two - three days late will incur a 5% loss. After three days, the loss will jump significantly to 25% or more. In unusual circumstances an extension may be granted, but only if you speak to your instructor in advance.
Penalties for a first violation of the Honor System in this course may include receiving no credit for an assignment, a lowered course grade, or failure of the course. Depending on the severity of the incident, a report may be sent to the Dean's Office, which may result in additional consequences, including suspension from the College. Any subsequent violation will result in immediate failure of the course.