Surrey University Stag

APPLIED MACHINE LEARNING - 2022/3

Module code: EEEM068

Module Overview

Machine/Deep learning has emerged from computer science and artificial intelligence. It draws on methods from a variety of related subjects including statistics, applied mathematics and more specialized fields, such as pattern recognition and neural network computation. This module offers the theory and related applications of advanced deep/machine learning topics and an overview their applications to other fields, such as natural language processing, medical imaging, health, audio, and fintech etc. The deep learning algorithms which will be studied are used widely in industry by AI start-ups to AI tech giants, like, Google, Meta, Microsoft, Amazon, Tesla etc. It provides a background and related theory of deep/machine learning to manipulate data from various domains like image, video, text, audio etc. This is done by various machine learning algorithms that are discussed, implemented, and demonstrated within the module.

Module provider

Electrical and Electronic Engineering

Module Leader

RANA Muhammad (Elec Elec En)

Number of Credits: 15

ECTS Credits: 7.5

Framework: FHEQ Level 7

JACs code:

Module cap (Maximum number of students): N/A

Overall student workload

Independent Learning Hours: 107

Lecture Hours: 33

Laboratory Hours: 10

Module Availability

Semester 2

Prerequisites / Co-requisites

None

Module content

Convolutional neural network (basic operations, separable convolution; skip connection)
Graph Convolutional neural network (GCNN)
Attention
Transformers
MLP
Graph attention network
Zero-shot/few-shot learning
Domain generalisation/transfer
Self-supervised learning
Multimodal analysis

Applications:
NLP
Computer vision
Fintech/fintech documents
Audio,
Medical imagaing

Assessment pattern

Assessment type Unit of assessment Weighting
Coursework Coursework 25
Examination Online Examination ONLINE (OPEN BOOK) EXAM WITHIN 4HR WINDOW 75

Alternative Assessment

NULL">

Assessment Strategy

The assessment strategy for this module is designed to provide students with the opportunity to demonstrate the skills learned within the module.

Knowledge of various machine learning theories and their applications, understanding of the engineering methodologies for design and implementation of software algorithms.

Knowledge of state-of-the art solutions to different machine learning and data analysis problems, their limitations and requirements for better solutions.

Skills of identifying, classifying and describing the performance of software systems and components through the use of analytical methods and modeling techniques.

Programming skills and understanding of software tools, development environments, libraries and reusable components such as Python, numpy, scikit-learn, PyTorch etc libraries. These software tools will be used to perform data processing and analysis, implementation of different machine learning algorithms.

Summative assessment and feedback

The summative assessment for this module consists of the following:

Examination : 60%

Coursework : 40% (weeks 5-10)

The examination will consist of 2h closed-book written examination. There will be 5 questions each from different area of the course. Each question consists of several subquestions testing knowledge, analytical, and design skills.

The coursework assessments will consist of several questions asking to implement a snippet of code performing different tasks related to data processing and analysis, pattern recognition and classification. The purpose of the coursework is for students to practically apply the theoretical knowledge gained from the lectures on a topic from each of the core themes of this module.

Formative assessment and feedback

For the module, students will receive formative assessment/feedback in the following ways.

¿ During lectures, by question and answer sessions
¿ During tutorials/tutorial classes
¿ During supervised computer laboratory sessions
¿ During participation of online quizes

Module aims

  • The aim of this module is to offer an in depth understanding of modern deep neural networks and associated applications. These modern AI/deep learning algorithms are universal and can be applied to many application areas, like, computer vision, robotocs, natural language processing, security & surveillance, medical image analysis, multimodal analysis, content retrieval of large multimeida archives, fintech etc. The goal is to make students ready to tackle real world problems by applying advanced deep learning/AI algorithms.

Learning outcomes

Attributes Developed
001 Be able to demonstrate an understanding of different principles of machine learning and data analysis. T
002 Be able to choose an appropriate method to a problem of computer vision, medical imaging, natural language processing, fintech etc and predict the outcome of the applied method. CKP
003 Be able to come out with a theoretical solution for the computer vision, medical imaging, natural language processing, and fintech problems. CKT
004 Be able to analyse novel pattern recognition and data analysis problems, establish statistical models for their solution. CP

Attributes Developed

C - Cognitive/analytical

K - Subject knowledge

T - Transferable skills

P - Professional/Practical skills

Methods of Teaching / Learning

The learning and teaching strategy is designed to achieve the following aims.
¿ To introduce advance deep learning concepts and their applications to promote deep understanding of the advanced AI concepts, engineering methodologies for design and implementation of software algorithms. The topics include: convolutional neural network, seperable convolutions, graph CNNs, attention, transformers, MLP, graph attention network, domain transfer/generalisation, zero/few-shot learning, self-supervised learning, multimodal analysis.
¿ To introduces engineering principles for software design in advance machine/deep learning, which involves integration of software components related to verification and validation.
¿ Students learn to identify, classify and describe the performance of software systems and components through the use of analytical methods and modelling techniques. These allow them to employ performance trade-offs and the use of appropriate metrics, to predict and evaluate performance of advance machine/deep learning algorithms.
¿ Students learn skills of applying quantitative methods, mathematical and computer-based models, and use computer software (Python, Pytorch) to solve image-processing problems. Skills involve identifying and analyse problems in advance machine/deep learning, and develop software solutions to adress problems in different application areas of machine/deep learning, for example, healthcare, security and entertainment.
¿ Students learn programming skills and understanding of software tools, development environments, libraries and reusable components such as python, pytorch and other machine/deep learning libraries. They use these to perform classification, segmentation, localisation, regression and other advance machine/deep learning tasks such as self-supervised learning and multimodal analysis.

We will finalise the aims after finalising the module contents so that things are in prespective

Indicated Lecture Hours (which may also include seminars, tutorials, workshops and other contact time) are approximate and may include in-class tests where one or more of these are an assessment on the module. In-class tests are scheduled/organised separately to taught content and will be published on to student personal timetables, where they apply to taken modules, as soon as they are finalised by central administration. This will usually be after the initial publication of the teaching timetable for the relevant semester.

Reading list

https://readinglists.surrey.ac.uk
Upon accessing the reading list, please search for the module using the module code: EEEM068

Other information

1. Aston Zhang, Zack C. Lipton, Mu Li, Alex J. Smola, Dive into Deep Learning, URL: https://d2l.ai 2. Ian Goodfellow and Yoshua Bengio and Aaron Courville, Deep Learning, MIT Press, 2016. URL: https://www.deeplearningbook.org/ 3. Michael M. Bronstein, Joan Bruna, Taco Cohen, Petar Veli¿kovi¿, Geometric Deep Learning Grids, Groups, Graphs, Geodesics, and Gauges, 2021. URL: https://arxiv.org/abs/2104.13478 4. Christopher M. Bishop, Pattern Recognition and Machine Learning, Springer, 2007. URL: https://www.microsoft.com/en-us/research/uploads/prod/2006/01/Bishop-Pattern-Recognition-and-Machine-Learning-2006.pdf

Programmes this module appears in

Programme Semester Classification Qualifying conditions
Computer Vision, Robotics and Machine Learning MSc 2 Compulsory A weighted aggregate mark of 50% is required to pass the module
Artificial Intelligence MSc 2 Compulsory A weighted aggregate mark of 50% is required to pass the module

Please note that the information detailed within this record is accurate at the time of publishing and may be subject to change. This record contains information for the most up to date version of the programme / module for the 2022/3 academic year.