Page 1 of 1

Module Code - Title:

CS5004 - DEEP LEARNING

Year Last Offered:

2021/2

Hours Per Week:

Lecture

2

Lab

0

Tutorial

2

Other

2

Private

4

Credits

6

Grading Type:

N

Prerequisite Modules:

Rationale and Purpose of the Module:

The objective of this module is to equip students with the fundamental knowledge and techniques necessary to effectively apply deep learning architectures using relevant data processing techniques and modelling approaches for real-world problems. Emphasis will be placed on training the student in the practical use of deep neural networks and deep learning platforms in order to create and deploy solutions.

Syllabus:

1. Deep learning workflow: data preparation, deep neural network model creation, model training, model optimisation and deployment. 2. An overview of deep neural network architectures - Convolutional Neural Networks, Recurrent Neural Networks, Long-Short-Term Memory models, Recursive Neural Networks, Generative Adversarial Networks). 3. Deep learning workflow on deep learning frameworks using a selection from Keras, TensorFlow, PyTorch, Caffe2, and MatLab. 4. Deep learning for computer vision - image segmentation, image analysis, image classification, image generation, object detection and tracking for video analysis. 5. Sample applications of deep learning across a variety of domains ranging from natural language processing to financial data analysis. 6. Emerging topics in deep learning such as Deep Reinforcement Learning.

Learning Outcomes:

Cognitive (Knowledge, Understanding, Application, Analysis, Evaluation, Synthesis)

On successful completion of this module, students should be able to: 1. Process and analyse data when furnished with large and real-world datasets for deep neural networks. 2. Select an appropriate deep learning architecture from a range of alternatives, and then apply it to a given real-world problem. 3. Demonstrate competency in the use of supported deep learning frameworks on a cloud hosted platform. 4. Interpret and evaluate the outputs of a deep learning architecture. 5. Compare and contrast deep learning architectures.

Affective (Attitudes and Values)

1. Given problems and data to investigate, the student will identify and discuss any significant ethical issues such as privacy, confidentiality, ownership, transparency and identity. 2. Given datasets, the student will question and demonstrate whether the data is representative and identify potential biases. 3. Following exposure to various frameworks and hosted platforms, the student will judge and challenge the limitations of current deep learning techniques.

Psychomotor (Physical Skills)

N/A

How the Module will be Taught and what will be the Learning Experiences of the Students:

The module will be delivered using a blended learning approach using on-line lectures, labs and tutorials.

Research Findings Incorporated in to the Syllabus (If Relevant):

Prime Texts:

Goodfellow, I., Bengio, Y., and Courtville, A. (2016) Deep Learning , MIT Press

Other Relevant Texts:

Chollet (2017) Deep Learning with Python , Manning Publications
Kim (2017) MATLAB Deep Learning: With Machine Learning, Neural Networks and Artificial Intelligence , Apress
Subramanian (2018) Deep Leaning with PyTorch , Packt Publications

Programme(s) in which this Module is Offered:

Semester - Year to be First Offered:

Module Leader:

j.j.collins@ul.ie