module specification

This module surveys essential principles, methods, and techniques in AI and machine learning. It covers a broad range of AI topics such as problem solving, knowledge representation, logical and probabilistic inference, and machine learning using methods of automata theory, logics, probability theory and statistics. It discusses examples of intelligent systems and studies how to develop applications that can learn from experience such as expert systems, automatic classifiers and autonomous systems planning their actions and communicating in natural language. Students will be offered lectures, which introduce key concepts, explain main principles and techniques in AI, and demonstrate how to apply them in areas such as image recognition and price forecasting.

The workshop will provide practical sessions to help students understand the content of the lectures and build the necessary skills to develop AI-applications using suitable problem descriptions and datasets.

The main topics of the study will include:

1. Introduction to Artificial Intelligence and Machine Learning. Rational reasoning and intelligent behaviour. Human and Artificial Intelligence. Examples of AI systems. (LO1)

2. Rational Agents. Conceptual model of rational agents. Environment types and agents classification. (LO2, LO3)

3. Problem Solving in State Space. Uninformed search and Heuristic Search. Comparison of the search algorithms. (LO2, LO3)

4. Knowledge Representation and logics. Facts, heuristics, assumptions, conditions, and conclusions. Logical modelling and logical inference. Logical theories and ontologies. (LO2, LO3)

5. Actions and Knowledge-based Planning. Linear, hierarchical and non-linear planning. Continuous planning and replanning. (LO2, LO3)

6. Uncertainty and Probabilistic Reasoning. Decision making under uncertainty. Probabilistic space and degree of truth. Probabilistic distributions. Conditional, prior and posterior probability. Bayes Rule. (LO2, LO3)

7. Learning Agents. Conceptual model. Supervised and unsupervised learning. Learning from observations. Decision tees, space of hypotheses and performance measurement. (LO2, LO3)

8. Statistical Learning. Classification, approximation and prediction. Probabilistic inference. Parameters learning using Bayes nets. Linear regression. (LO2, LO3)

9. Neural Networks. Computation in the brain and artificial neuron models. Linear neural networks, multi-layer networks. Backpropagation. Application to image processing. (LO2, LO3)

10. Natural Language Processing. Communication and language. Language structure and linguistic data. Parsing and syntactic processing of written texts. Problems. (LO2, LO3)

11. Legal, Social, Ethical and Professional Issues in AI. AI and explanation. AI and law. Ethics of AI and impact of AI on society. (LO4)

The module is taught by a mixture of lectures, tutorials, workshops, and self-study activities.

•The lectures (1hour) are normally used to introduce key concepts, principles, models and algorithms related to AI and machine learning. The lectures can be recorded for self-study.

•Each lecture is followed by a 2-hourworkshop session. In these sessions the students will solve problems requiring AI to ensure they understand key concepts and methods and have some insight into the models and algorithms. They will be given some Python code related to the lectures and will be required to expand and modify it in order to produce and report the solutions.

•Students are expected to spend time for a) revising the lectures and preparing for the final test, completing and expanding the workshop tasks and c) unsupervised work on their coursework in the computer laboratories and in private.

All lecture and workshop materials for students will be placed on WebLearn. Past exam papers with or without sample answers could be provided as an additional material to help with the preparation for the final exam.

On completing the module, the student will be able to:

- LO1: Understand and critique the principles of rational reasoning and intelligent behaviour, the similarities and the differences between natural and human intelligence.

- LO2 Formulate problems which require AI approach for solving them and choose appropriate agent architectures

- LO3: Apply knowledge of the most popular models, methods and algorithms for building and operation of rational agents and for processing information using AI

- LO4: Design and develop basic AI programs which demonstrate intelligent behaviour and rational thinking in typical environment using available data sets.

- LO5: Understand the Legal, Ethical & Professional Issues brought by AI and comment on the impact of AI on individuals, organisations and the society as a whole.

This is the e-reading list link:

https://rl.talis.com/3/londonmet/lists/1921F050-E635-9C5B-9CCE-82F361938484.html?draft=1&lang=en-GB&login=1

Textbooks

• Stuart J. Russell, Peter Norvig. Artificial intelligence: a modern approach. 4th edition, Pearson, 2022 (Core)
• David L. Poole, Alan K. Mackworth. Artificial Intelligence: Foundations of Computational Agents. 3rd edition. Cambridge University Press, 2023 (Optional)
• Andrij Burkov. The Hundred-Page Machine Learning Book. Amazon Fulfillment, 2019 (Optional)

Programming guides

• Jim Smith. Learn Artificial Intelligence With Python: Learn How to Program Artificial Intelligence in One Day and Learn it Well. Independent publishers, 2023 (optional)
• Andreas C. Müller, Sarah Guido. Introduction to machine learning with Python: a guide for data scientists. O'Reilly (2017) (Optional)
• Enes Bilgin. Mastering Reinforcement Learning with Python: Build next-generation, self-learning models using reinforcement learning. Packt Publishing, 2020 (optional)