AI & Machine Learning (2024 SPRING)
*** REMINDER ***
This course is an elective course particularly designed for the ITM MBA students specializing in Electronic Commerce.
Students are expected to attend all lectures and complete all courseworks.
Attendance will not be accounted for a part of the final score.
But, absence without proper reason will be.
Professor John Sum is a strict person and sometimes a monster. His expectation on the behavior and the performance of a student is stringent.
It is simply because John Sum has been teaching in the area of intellient technology for more than three decades.
Materials in his mind are abundant. Only some of them are extracted and complied as teaching materials.
*** REMINDER END ***
Students are expected to have the following skills and knowledge.
- Students should have elementary level of English and have no problem in reading English materials.
- Skillful in the use of word processing software for report writing and presentation.
- Skillful in searching information over the Internet.
- Knowledge in one of the following subjects:
Principles of Computing, Computer Literacy, Introduction to Information Systems, Introduction to Computer Science,
Introduction to Information Technologies or other related subjects.
(To refresh the concepts in 'Introduction to Computer Science',
here is a link to the course 'Introduction to Computer Science'
which was delivered by me in the 2022 Fall term for the BBA students in our management school.)
- Able to complete a coursework report as compared with a jorunal publication.
Students are expected to have a good health. Then, the students are able to work over serval nights.
Each student needs to have a valid email account for assignment submission, written report submission and communications with the instructor.
Here lists the tentative topics to be introudced and discussed in the course.
- Introduction to Intelligent Technology (JS).
- Introducing Generative AI.
- Large language models: OpenAI ChatGPT, Google Bard/Gemini.
- Chinese LLMs: Baidu Ernie, Tencent Hunyuan,
BAAI WuDao.
- Offline LLM models: LLM services without Internet connection.
- TechEmergent (2023), 60 best generative AI companies and startups in 2023.
- Midjourney: A text-to-image generator.
- Question: What is(are) the possible application(s) of text-to-image generation?
- Create an Internet celebrity using free AI tools, like ChatGPT, Google Translate, Midjourney and DeepFake.
- YouTuber experience sharing: Towards a profitable YouTuber with the use of AI tools.
- Microsoft Copilot: Your everyday AI companion. A collection of generative AI models for use.
- Theory behind generative AI models: Latent class models and Structural equation models.
- Goodfellow, I. et al. (2014). Generative adversarial networks. arXiv preprint, arXiv:1406.2661.
- Goodfellow, I. et al. (2020). Generative adversarial networks. Communications of the ACM, 63(11), 139-144.
- Radford, A., Metz, L., & Chintala, S. (2015). Unsupervised representation learning with deep convolutional generative adversarial networks. arXiv preprint, arXiv:1511.06434.
- Wikipedia, Factor analysis: A generative model for statistical analysis in social science research.
- Technologies for Information/Intelligent Infrastructue.
- Intelligent Products and Services (JS).
- Essentials of Intelligent Technology (JS).
- Technologies for language understanding.
- Technologies for image recognition.
- Core AI/ML technologies.
- Federated learning (Collabrative learning).
- Intelligent Services Development (JS).
- Intelligent technology development (hard) = AI model development = Foundation model development.
- Theoretical analysis.
- Empirical analysis.
- Software/Hardware implementation.
- Hardware design, i.e. processor design.
- Intelligent service development for fun.
- Intelligent service development for profit = Application system development.
- Intelligent Technology for Manufacturing.
- Manufacturing execution systems (beware of the terminology).
- Digitial twin.
- Don't Expect Too Much.
- AI + Digital Twin.
- With digital twin, a light-off factory can now be visualized and controlled (remotely) from its digital twin.
- With AI, the light-off factory could be made to be an autonomous system.
- Predicting and visualizing the coming behavior of a system.
- Note: A digital twin can simply be a bunch of data showing the current status of a system.
- AI algorithms for solving optimization problems are not really intelligence.
- Questions.
- What intelligent technologies have been applied?
- How does AI help in solving the optimization problems in manufacturing?
- How does AI help in new product design?
- AI in Manufacturing: China Experiences. (3+ hours)
- Practical AI use cases for manufacturing, by Siemen and Maya HTT (54 minutes).
- Advanced Topics on Intelligent Technology Management.
Use of AI Tools
AI tools, like ChatGPT and Google Bard, have to be used with extrememly careful.
Some AI tools might give incomplete information or non-logical arguments in response to a question (equivalently, a prompt).
Moreover, Google Bard does not provide any reference for the content generated.
So, a student has to validate if the information conveyed in the sentences is completed and logical.
Owing to avoid plagiarism, a student has nevertheless to justify if the sentences resemble some sentences already appeared in some published articles.
ChatGPT, Google Bard and other AI text generators are very good in paraphrasing.
Using these tools to paraphrase your assignments and project reports are highly recommended.
Tips:
In the first step, you have to complete your assignment or the project report based upon your own writing.
In the second step, you could use those AI tools to paraphrase the contents of your assignments and reports.
WARNING: For any part of the content (respectively, any word) you have put in an assignment or report,
I will ask for your reason and explanation why this part of content (respectively, the specific word) has to be added.
If you fail to do so, your assignment (respectively, group project) score will be zero.
Nevertheless, your assignment (respectively, group project) score will be zero if you fail to give the source for any part of the content you have added
in your assignment (respectively, group project).