CS7003 - Advanced AI Technologies (2024/25)
| Module specification | Module approved to run in 2024/25 | ||||||||||||
| Module title | Advanced AI Technologies | ||||||||||||
| Module level | Masters (07) | ||||||||||||
| Credit rating for module | 20 | ||||||||||||
| School | School of Computing and Digital Media | ||||||||||||
| Total study hours | 200 | ||||||||||||
|
|||||||||||||
| Assessment components |
|
||||||||||||
| Running in 2024/25(Please note that module timeslots are subject to change) |
|
Module summary
Artificial Intelligence research and development spans over a period of nearly 70 years. During this period the academicians have been trying to address most of the aspects of human intelligence known from other fields – mathematics, philosophy, psychology, linguistics, biology, etc. As a result, AI was clearly partitioned into several areas, each with its own methodology of investigation and technology of problem solving – state-space problem solving, decision making, automated reasoning, knowledge-based planning and machine learning. While during different periods one or another were attracting the attention, all of them found practical applications and gradually evolved, reaching bigger depth and maturity.
This module covers the evolution of AI and provides a thematic coverage of the various contemporary branches of the discipline. AI paradigms, models, and technologies are covered and students are given the opportunity to investigate and apply those that interest them most.
Syllabus
Theme 1: Advanced vs. classical AI. Linguistic and logical models, Symbolic and computational methods, Knowledge-based and Data-focused systems. Linguistic and Visual applications. Challenges they present from ethical, legal, psychological and social point of view [LO1, LO3]
Theme 2: Advanced Problem-solving Models and Methods. Bayesian Networks. Markov Chains and Hidden Markov Models (HMM) [LO2-LO5]
Theme 3: Advanced Decision-making Models and Methods. Markovian Models and Reinforcement Learning. Deep Reinforcement Learning [LO2-LO5]
Theme 4: Advanced Machine Learning. Deep Learning and Transfer Learning. Convolution Neural Networks (CNN) and Recurrent Neural Networks (RNN). [LO2-LO5]
Theme 5: Advanced Knowledge-based Systems. Description Logic, Ontological Modelling and Serialization. Graph Databases and Knowledge Graphs. Semantic Information Processing [LO2-LO5]
Theme 6: Advanced Linguistic Systems. Sentiment Analysis and Deep Learning for NLP. Transformers and Large Linguistic Models (LLM). ChatGPT [LO2-LO5]
Balance of independent study and scheduled teaching activity
The module combines independent study for researching the philosophical, ethical, social and legal aspects of AI with scheduled teaching activities to learn the fundamental knowledge and the practical skills needed to build intelligent systems using advanced AI models, methods and technologies.
The blended learning is organised by combining different options:
learning resources: lecture slides, academic textbooks, public standards, workshop tasks, software tools, coursework case studies, software documentations, online discussion forums
teaching methods: lecturing, group tutoring and individual supervising, group presentation and individual demonstration
working modes: independent study, collaborative and supervised design and development, public presentation and demonstration
assessment mechanisms: individual and group assignments
Learning outcomes
LO1 Understand the differences between classical and advanced problems, paradigms and methodologies in AI and their challenges from ethical, legal, psychological and social point of view
LO2 Learn advanced methods and algorithms for modelling of intelligent reasoning and behaviour
LO3 Develop some interest and ability to do independent study of more complex models, more sophisticated methods and more complex technologies
LO4 Practice modelling of intelligent applications which utilize advanced AI models and methods
LO5 Acquire practical skills for design and development of AI systems which use advanced AI technologies
Bibliography
Core Textbooks
[1] Peter Norvig and Stuart Russell. Artificial Intelligence: A Modern Approach, 4th edition, Pearson (2021); ISBN-13: 978-1292401133
[2] Grigoris Antoniou, Paul Groth, and Frank Van Harmelen. A Semantic Web Primer, 3rd edition, MIT Press (2012); ISBN-13: 978-0262018289
Additional Books
[3] Mayank Kejriwal, Craig Knoblock, and Pedro Szekely. Knowledge Graphs: Fundamentals, Techniques, and Applications, The MIT Press (2021); ISBN-13: 978-0262045094
[4] Daphne Koller and Nir Friedman. Probabilistic Graphical Models: Principles and Techniques, The MIT Press (2009); ISBN-13: 978-0262013192
[5] John Kelleher. Deep Learning, MIT Press (2019); ISBN-13: 978-0262537551
[6] Richard Sutton and Andrew Barto. Reinforcement Learning: An Introduction, 2nd edition, MIT Press (2018); ISBN-13: 978-0262039246
[7] Lewis Tunstall, Leandro von Werra, and Thomas Wolf. Natural Language Processing with Transformers: Building Language Applications With Hugging Face, O'Reilly Media (2022); ISBN-13: 978-1098136796
Free Online Resources
https://github.com/protegeproject
Stanford Ontological Editor Protégé
https://graphdb.ontotext.com/
Ontotext Graph Database GraphDB