Pattern Clasification has shifted the traditional delivery of lectures into a multi-faceted teaching methodology that includes classroom teaching, project development, in-depth discussions and seminars on various AI projects.
According to the plan, this unit needs a total of 48 class hours, including 32 hours teaching, and 16 hours student presentation/discussion. The unit content is structured around the following key aspects:
-
Building a strong foundation: The unit focuses on providing fundamental knowledge of pattern recognition, including Bayesian decision theory, MLE/MAP estimation, non-parametric methods, dimensionality reduction and stochastic methods. By gaining a thorough understanding of these fundamental concepts and principles, students will be equipped to undertake real-world pattern recognition projects.
-
Developing engineering skills: Students are encouraged, inspired and guided to choose their own topics or to participate in hands-on research conducted by unit team members. This approach enables students to apply their acquired knowledge and enhance their creative skills. The unit places a strong emphasis on practical skills and encourages student interaction, collaboration, independent thinking and problem solving.
-
Instilling values and cultivating a scientific mindset: The unit aims to instil in students the spirit of a scientist. Theoretical courses embed this scientific spirit, while practical courses emphasise its application. By fostering qualities such as the pursuit of truth, kindness, humility, critical thinking, innovation and cooperation, the unit contributes to the development of students' scientific character. It encourages questioning, analytical thinking, hypothesising, testing, experimenting and perseverance.
-
Exploring scientific frontiers: The unit emphasises the integration of knowledge with the latest advances in the field. At the end of each knowledge module, the latest research in the field is presented, accompanied by a comprehensive bibliography. This provides students with the necessary skills to explore and delve into further study independently. The course also links theoretical knowledge to practical problems and national development strategies, thus fostering a sense of national responsibility among students.
In this unit, all lecture slides handouts will be provided prior to the class, so that students can prepare in advance and have some question prepared for the lecture sessions. Moreover, the practicals are all available publicly, and students are expected to work on those practicals themselves (especially for deliveries which have no practical allocation).
Just because we give you a lot of the content in advance, it doesn't mean that we expect you to teach yourself. You do the learning – we can't do that part instead of you – but you can still leave the teaching to us: you will have ample opportunities to ask questions in the class: interactive activities will be done in the weekly lectures and/or practical classes (if with time allocation). For those who can't make it to the live sessions or have urgent questions that can't wait for the next available class, through the discussion forums at GitHub
To facilitate the study, we have organised the contents into a structure where each session corresponding to one set of classes. In each session, you are provided with the lecture slides handouts, notes as well as practical activities.
Within this unit, you'll find preliminary sessions, core sessions and advanced sessions. We'll explain each of these, and how they relate to your unit assessment.
We would like you to go through the sessions at the time of your ease, and then work on the assessment tasks. The first thing you'll notice when you begin working through the sessions is that each has a consistent format, especially featuring:
- A
sessionREADME.mdpage: here you'll find the overview for the session, and most download materials are available through this page. It is especially important that you make note of these as they explain the sessions structure and the supplementary resources for the session. In order to successfully complete each session, you will need to ensure that you have achieved each of these objectives.
We assume no prior Python experience, and any prior knowledge on programming will be a bonus: if you already have completed several units such as machine learing, deep learning, you should be reasonably confident about your Python programming skills.
In case that you have no Python experience, SIT742 - 📖 Python Foundations will prepare you in this aspect and enable you to continue into the core sessions.
At the heart of the unit are six core sessions. These core sessions address the statistical pattern recognition techniques that are relevant and important for AI researchers and data analysters. Therefore, it is an essential requirement of this unit that all core sessions are completed.
In additional to the core sessions, there are also a series of advanced topics. These topics are optional and they are either build on the information in the core sessions by exploring concepts in greater depth, or introduce new, more specialised topics.
They are there for those wishing to expand their knowledge or explore specific topics in further detail. Depending on your course, you may like to choose those advanced topics which are most relevant.
Develop knowledge of pattern recognition in order to select or develop appropriate algorithms and models for given applications; (GLO1, GLO2, GLO3)
Use different measures to evaluate the costs and risks associated with different decisions and to identify optimal solutions to specific pattern recognition problems; (GLO4, GLO5)
Apply pattern recognition techniques in a real-world project, work collaboratively in a team to develop the solution and present the methodology and results for a given scenario; (GLO5, GLO9, GLO10)
Different universities have different set of GLOs, for example:
👉 GLO for Southeast University
1️⃣ Engineering knowledge: a solid mathematics required in computer engineering, natural sciences, engineering base and expertise to solve complex engineering problems.
2️⃣ Problem Analysis: able to apply the basic principles of mathematics, natural sciences and engineering sciences to identify, express, and through literature research and analysis of complex engineering computer engineering problem to reach a valid conclusion.
3️⃣ Design / development solutions: can integrate the use of theory and techniques to design solutions to complex engineering problems in the field of computer engineering, designed to meet the information access, transmission systems, processing or use of other needs, the unit (member) or process processes, and to reflect the sense of innovation in the design session, taking into account the social, health, safety, legal, cultural and environmental factors.
4️⃣ Research: it can be based on scientific principles and scientific methods for complex engineering problems in computer engineering field studies, including design of experiments, analysis and interpretation of data, and through comprehensive information reasonably valid conclusions.
5️⃣ Use of modern tools: the ability to solve complex engineering problems in computer engineering, to develop, select and use appropriate technology, resources and modern engineering and information technology tools, including prediction and simulation of complex engineering problems in computer engineering and the ability to understand their limitations.
...
9️⃣ Individual and Teamwork: the ability to work individually, as a team member and as a responsible member of a multidisciplinary team.
🔟 Communication: the ability to effectively communicate and exchange complex engineering computer engineering problems with industry peers and the public, including writing reports and design documents, statements speak, articulate or respond to commands. And have some international vision, able to communicate and exchange in a cross-cultural context.
We aim to generate enthusiasm for pattern recognition and improve teaching effectiveness through lecturer experience sharing, session-based discussion and real-world project-based assessment.