COM1005 Machines and Intelligence
|| This module provides an introduction to Artificial
Intelligence, and to key concepts and problems in the field,
such as whether a computer is capable of understanding, and
whether humans should themselves be viewed as machines. It
also provides a brief historical overview of the subject and
reviews the state-of-the-art and open questions in some of
the major sub-areas of AI, pointing out connections to
research work in the Department. As well as providing a
first encounter with the main issues that underlie attempts
to create Artificial Intelligence, the module also has a
more practical component that introduces algorithms and
data structures for AI problem solving through practical
programming examples, as well as hands-on experience with simple programming of robots. The emphasis here is on identifying
the abstract nature of the problem which is to be solved,
matching this to an appropriate algorithm or technique and
implementing a solution. It also serves as an introduction
to programming for research rather than for software
|| Academic Year 2021/22
- Blackboard quizzes
- Group assignment
|| Prof. Rob Gaizauskas, Prof. Tony Prescott & Dr Heidi Christensen
|| The aims of the module are
- to provide students with the historical and cultural
context of modern day research into artificial
- to introduce the student to the AI research carried
out in the Department.
- to introduce a number of classic AI problem-solving
algorithms and data structures
- to develop an ability to select appropriate
techniques to address particular problems.
- to develop the technical knowledge necessary to
implement AI problem solving
- to provide experience of scientific programming as
opposed to software engineering.
|| By the end of the course the students should be able to:
- Discuss the main issues involved in defining intelligence
- Explain representative AI programs that are introduced in the module
- Explore how AI is being used to create autonomous agents such as robots
- Demonstrate the ability to select and implement appropriate techniques to address particular problems
- Demonstrate the application of simple AI programs in robotics
||Semester 1 content:
Semester 1 addresses foundational questions in AI about
the nature and possibility of artificial intelligence,
provides a brief historical overview of the subject and
also reviews the state-of-the-art and open questions in
some of the major sub-areas of AI.
Students will use both physical and simulated robots as an introduction to programming for research.
Topics covered in Semester 1:
- What is intelligence? Is AI Possible?
- Turing Test
- Chinese Room Argument and Strong versus Weak AI
- Physical Symbol System Hypothesis and
- Origins and Early History of AI
- Open Questions in some of:
- Machine Learning
- Machine Vision
- Speech and Language Understanding
- Knowledge Representation and Reasoning
- Ethics and Artificial Intelligence
Semester 2 content:
In contrast to semester 1, the objective here is to give
students knowledge and experience of AI programming by
implementing classic symbolic problem-solving paradigms.
It also provides students with experience of programming
for scientific research in contrast with software
Topics covered in Semester 2:
- State-space search: search algorithms and strategies.
- The dynamic programming principle and A* search
- Symbolic knowledge representation and pattern
- Reasoning in deductive systems: forward and backward
chaining, production systems, rule networks and the RETE
- Planning and goal reduction
- AI, Connectionism and Perception - limitations of symbolic reasoning.
|| Lecture based, with assignments and lab classes in semester 1. Assignments and lab classes in semester 2.
|| Semester 1: Assignments will be marked using published criteria and returned within 3 weeks of submission. General feedback given during face to face labs and online interactive sessions.
Semester 2: Feedback on labs during the sessions as well as in lectures. Assessments will be marked using published
criteria and the submissions commented. They will be
returned within 3 weeks of submission.
- Russell and Norvig: Artificial Intelligence: A Modern
Approach [it is not necessary to buy this -- there is an e-copy in the library]
- Joe Minichino, Joseph Howse. Learning OpenCV 4 Computer Vision with Python 3: Get to grips with tools, techniques, and algorithms for computer vision and machine learning, 3rd Edition.