ECE 677

Course Description

As one of the cornerstones of modern artificial intelligence, neural networks have revolutionized industries ranging from image recognition and natural language processing to self-driving cars and healthcare diagnostics. The course objective is to prepare students to learn deep learning and neural networks techniques using contemporary open-source tool sets. The course starts with the fundamental building blocks of neural networks and progresses to advanced topics, ensuring a strong grasp of both theory and practical implementation. Students will apply the acquired skills through multiple artificial intelligence related engineering applications. Basic concepts needed for deep learning development such as linear algebra and calculus will be covered.

Learning Outcomes

1. Master Neural Network Fundamentals by grasping core concepts, neuron functionality, and architecture essentials.

2. Design MLPs, CNNs, RNNs, and transformers for various tasks.

3. Develop networks with backpropagation, optimization, and regularization.

4. Employ TensorFlow and PyTorch for real-world data.

5. Improve models through tuning, transfer learning, and regularization.

6. Solve image analysis, NLP, and other challenges using tailored networks.

7. Gain insights into advanced topics such as GANs, reinforcement learning, and model interpretability.

8. Address biases and deploy AI responsibly.

9. Stay updated on evolving trends and technologies.

10. Convey ideas and results clearly for collaboration and impact.

Textbooks

• Lecture Notes

• "Deep Learning" Ian Goodfellow, Yoshua Bengio, and Aaron Courville, MIT Press, 2016

Schedule