Syllabus

Syllabus#

Course Description#

AINS6003 Deep Learning & Neural Networks explores neural network architectures—including multilayer perceptrons, convolutional, recurrent, and transformer models—with emphasis on training techniques, optimization, regularization, and applications in vision, speech, and generative AI. The course aligns with Aurnova AIN6003 (3 credits, online, 8-week block).

Prerequisites#

  • AINS6002 Machine Learning & Predictive Modeling (or equivalent)

  • Comfort with Python, linear algebra, and basic probability

  • Access to GPU-backed compute for later modules (Codespace, Colab Pro, or institutional credits)

Learning Outcomes#

By the end of the course, students should be able to:

  • explain forward and backward propagation in feedforward networks

  • select optimizers, loss functions, and regularization for a given task

  • design CNN, RNN/LSTM, and transformer-based solutions at a conceptual and implementation level

  • compare discriminative and generative modeling approaches

  • profile and document GPU training workflows for reproducibility

Assessment Pattern#

Component

Weight

Module assignments (6–8 submissions)

50%

Lab notebooks (completion + reflection)

20%

Mid-course architecture design brief

15%

Final synthesis / capstone-style project

15%

Weekly rhythm#

Each module includes: overview, book prose, assignment, slides, narration, instructor notes, and a lab notebook.

Resources#

Schedule (example 8-week block)#

Week

Module

Theme

1

1

Neurons, MLPs, activations

2

2

Backpropagation and autograd

3

3

Optimization and regularization

4

4

Convolutional networks

5

5

Sequence models

6

6

Attention and transformers

7

7

Generative models

8

8

GPU workflows and deployment