About the subject
What Artificial Intelligence is for
The 2080 curriculum brought this course up to date. Alongside the classic foundations of AI, search algorithms like A* and minimax, logic and knowledge representation, it now includes machine learning with decision trees and support vector machines, neural networks and backpropagation, deep learning, and newer topics such as agentic AI, federated learning, edge AI and responsible AI.
The course is both theory and practice. Twelve laboratory sessions take students from search and game playing to a machine learning pipeline and neural networks, and end in a mini project. BEI students study artificial intelligence too, as ENCT 305 in Semester 5, with a different outline.
The objective of this course is to provide students with a foundation in Artificial Intelligence (AI), covering intelligent agents, search techniques, knowledge representation, machine learning, and AI ethics. It aims to equip students with both theoretical understanding and practical skills to apply AI techniques to real-world problems, while also developing awareness of the ethical and societal implications of AI systems.
IOE's course objective for ENCT 351
- Taught to
- BCT, Semester 6 (Year III, Part II)
- Weekly
- 3 lecture, 1 tutorial, 3 practical hours
- Marks
- Theory 40 internal + 60 final (3-hour exam); practical 50 internal; 150 in total
Where the marks are
ENCT 351 chapters, hours and final exam marks
IOE's evaluation scheme for the 60-mark final. IOE notes there may be minor deviation.
Scroll the table sideways for hours, marks and share.
Full syllabus
The complete ENCT 351 outline, with how to study each chapter
All 7 chapters and 41 topics as IOE lists them, each with ICE's advice on approaching it.
1Introduction4 hours
Foundations and history of AI, intelligent agents and their types, and agentic AI.
- 1.1 Definition, foundation, history of AI
- 1.2 Importance of knowledge and learning
- 1.3 Cognition and learning (Neuroscience)
- 1.4 Human intelligence and machine intelligence
- 1.5 AI tree: Branches and interdisciplinary nature
- 1.6 Intelligent agents and types
- 1.7 Agentic AI
2Problem Solving and Search9 hours
Search, 9 hours and 12 of 60 marks: uninformed and informed search, A*, minimax with alpha-beta pruning, hill climbing, simulated annealing and genetic algorithms. Trace each algorithm by hand on a small graph.
- 2.1 Formal problem definition: States, actions, transitions, well-defined problems
- 2.2 Constraint satisfaction problems: Node consistency, path consistency, backtracking
- 2.3 Search algorithms, strategies and evaluations
- 2.4 Uninformed: BFS, DFS, iterative deepening
- 2.5 Informed Search: Best first search, greedy search, A* algorithm
- 2.6 Adversarial search: Minimax algorithm, alpha-beta pruning
- 2.7 Local search and optimization: Hill climbing, simulated annealing
- 2.8 Evolutionary optimization: Genetic algorithm
3Knowledge Representation and Probabilistic Reasoning7 hours
Logic, semantic networks, frames and knowledge graphs, Bayesian reasoning and fuzzy inference, 10 marks.
- 3.1 Knowledge-based agent
- 3.2 Knowledge representation techniques and issues in representation
- 3.3 Propositional and predicate logic
- 3.4 Semantic networks, frames and knowledge graph
- 3.5 Review of Bayes' Theorem and probabilistic reasoning
- 3.6 Fuzzy logic: Membership functions, fuzzy inference systems
4Machine Learning Fundamentals10 hours
Machine learning fundamentals, 10 hours and 14 marks, the joint largest share: optimisation, the ML pipeline, decision trees, support vector machines, t-SNE, and semi-supervised and reinforcement learning. The mathematics from Probability and Statistics is used here.
- 4.1 Foundations and four pillars of machine learning
- 4.2 Review of mathematics for machine learning
- 4.3 Continuation optimization; Unconstraint optimization, constraint optimization, convex optimization
- 4.4 Review of machine learning pipeline
- Model development: Generative versus discriminative
- Learning algorithm
- Capacity, overfitting and underfitting
- Hyperparameters and validation sets
- Estimators, bias and variance
- Review of MLE and MAP and cross entropy
- 4.5 Supervised learning algorithm: Decision tree and support vector machine
- 4.6 Unsupervised learning algorithm: t-SNE
- 4.7 Semi supervised and reinforcement learning
- 4.8 Model evaluation: Energy-based indicators
5Neural Networks and Deep Learning Algorithms8 hours
Neural networks, perceptrons, backpropagation and an introduction to deep, recurrent and generative networks, 10 marks. Work one backpropagation example fully by hand.
- 5.1 Neural networks: Structures, activation functions and universal approximation theorem
- 5.2 Perceptron, multilayer perceptron and backpropagation
- 5.3 Introduction to deep learning
- 5.4 Concepts on recurrent and generative neural networks
6AI Applications5 hours
Expert systems, natural language processing, robotics and computer vision, and sustainable AI.
- 6.1 Expert systems: Characteristics, architecture, development and various applications
- 6.2 NLP: Level of analysis, challenges, modern approaches and applications
- 6.3 Robotics and computer vision: Fundamental, components and applications
- 6.4 Sustainable AI systems
7Emerging Trends2 hours
Sequence-to-sequence models, federated learning, edge AI and responsible AI. Chapters 1, 6 and 7 together carry 14 marks.
- 7.1 Sequence to sequence models
- 7.2 Federated learning
- 7.3 Edge AI
- 7.4 Ethics and AI: Responsible AI
Laboratory
The ENCT 351 practical
The practical is three hours a week with twelve sessions: agents and problem formulation, uninformed, informed and adversarial search, constraint satisfaction, genetic algorithms, knowledge representation, probabilistic and fuzzy reasoning, the machine learning pipeline, supervised and unsupervised learning, neural networks, and a mini project with applications and ethics. Implement the algorithms yourself before using a library.
- Intelligent agents and problem formulation
- Uninformed search techniques
- Informed (Heuristic) search techniques
- Adversarial search and game playing
- Constraint satisfaction problems (CSP)
- Evolutionary computation (Genetic algorithms)
- Knowledge representation (Logic, semantic networks, frames)
- Probabilistic and fuzzy reasoning
- Machine learning pipeline and data preprocessing
- Supervised and unsupervised learning
- Neural networks and deep learning basics
- Mini project, AI applications and ethics
Before and after
How Artificial Intelligence connects to other courses
Builds on
It closes Year III. The Year IV electives, whose syllabi IOE has not yet published, are where students can go further.
Reference books
Books IOE lists for ENCT 351
- Russell, S., Norvig, P. (2021). Artificial intelligence: A modern approach. Pearson.
- Rich, E., Knight, K., Nair, S. B. (2009). Artificial intelligence. McGraw-Hill Education.
- Bishop, C. M. (2006). Pattern recognition and machine learning. Springer.
- Deisenroth, M. P., Faisal, A. A., Ong, C. S. (2020). Mathematics for machine learning. Cambridge University Press.
- Goodfellow, I., Bengio, Y., Courville, A. (2016). Deep learning. MIT Press.
Quick answers
Artificial Intelligence questions
Is machine learning in the IOE Computer Engineering AI syllabus?
Yes. Chapter 4, Machine Learning Fundamentals, is the joint largest chapter with 14 of 60 marks, and chapter 5 covers neural networks and deep learning.
In which semester is Artificial Intelligence taught in BCT?
Semester 6, Year III Part II, as ENCT 351. BEI students take a separate AI course, ENCT 305, in Semester 5.
Does IOE AI include a project?
Yes. The last of the twelve laboratory sessions is a mini project on AI applications and ethics.
How many credits and marks is ENCT 351 Artificial Intelligence?
3 credits and 150 marks: 40 internal and 60 in a 3-hour IOE final for theory, plus 50 marks of practical assessed internally. It is taught 3 lecture, 1 tutorial and 3 practical hours a week.
Source
Checked against IOE
The outline, references and marks are IOE's own, from the ENCT 351 syllabus PDF and IOE's curriculum structure. The study advice is ICE's. If IOE revises the course, its syllabus is what counts. IOE's BCT curriculum page.
