We designed a 4-module sequence for an undergraduate deep learning course. Modules are designed to give students hands-on experience with deep learning frameworks and concepts through working with datasets of varying types. The coverage for modules includes:
Module 1) introduction to linear regression analysis by working on Facebook Metrics Dataset for modeling total user interactions. Students have also been introduced to non-linear datasets and how kernel methods and weight regularization can be used for non-linear modeling patterns.
Module 2) through working on the Fashion-MNIST dataset, students will 1) understand how to pre-process image data and conduct dimensionality reduction, 2) implement logistic regression from scratch, and 3) implement a neural network and observe the capacity of neural networks to learn a non-linearly separable decision boundary.
Module 3) focuses on designing convolutional neural networks and training them on CIFAR-10 dataset. The modules also introduce transfer learning.
Module 4) students will work on training recurrent neural networks on the Reuters newswire classification dataset and will analyze their observations.
This 4-module sequence is easily adoptable by instructors and is implemented in a modular structure. The target audience of these modules is undergraduate CS or engineering students taking AI or ML courses that cover topics such as 1) linear regression, 2) logistic regression, 3) neural networks, 4) convolutional neural networks, and 5) recurrent neural networks. The prerequisites for the course include programming, data structures, algorithms, and probability courses.
Assignments are designed to give students hands-on experience with deep learning frameworks and concepts through working with datasets of varying types.

Linear Regression
Logistic Regression
Neural Networks
Convolutional Neural Networks
Recurrent Neural Networks

The target audience of these assignments is undergraduate CS or engineering students taking AI or ML courses that cover topics such as:
1) linear regression, 2) logistic regression, 3) neural networks, 4) convolutional neural networks, and 5) recurrent neural networks.
The prerequisites for the course include programming, data structures, algorithms, and probability courses.

These assignments covers all steps of teaching machine learning and deep learning to beginners. Instructions in the handouts will make the learning experience smooth and engaging.
Instructors can pick and use different modules that were covered in this set of assignments depending on the level of the course they teach.