AI Product Management

 

Completed a specialisation relating to AI Product Management

Ever since I subscribed to the US$1 Coursera Subscription (for the 1st month only) in early December, I have been actively completing several hands-on projects, courses and clinching specialisations on the Coursera platform as part of my own professional development. I dived deeper into data analytics which I had blogged about in a previous post, and I tried to pick up some skills related to AI and Machine Learning by enrolling on a course by Duke University on Product Management and another on Machine Learning: Algorithms in the Real World Specialization by Amii (Alberta Machine Intelligence Institute). I managed to clear the former but am still in progress for the latter. For those who may be keen, I have included some details about the 1st course in the specialisation that I have completed below and I will also share some screenshots and reflections from my experience learning in this MOOC (Massive open online course).

Machine Learning Foundations for Product Managers

by Duke University

In this first course of the AI Product Management Specialization offered by Duke University's Pratt School of Engineering, you will build a foundational understanding of what machine learning is, how it works and when and why it is applied. To successfully manage an AI team or product and work collaboratively with data scientists, software engineers, and customers you need to understand the basics of machine learning technology. This course provides a non-coding introduction to machine learning, with focus on the process of developing models, ML model evaluation and interpretation, and the intuition behind common ML and deep learning algorithms. The course will conclude with a hands-on project in which you will have a chance to train and optimize a machine learning model on a simple real-world problem. At the conclusion of this course, you should be able to: 1) Explain how machine learning works and the types of machine learning 2) Describe the challenges of modeling and strategies to overcome them 3) Identify the primary algorithms used for common ML tasks and their use cases 4) Explain deep learning and its strengths and challenges relative to other forms of machine learning 5) Implement best practices in evaluating and interpreting ML models 

The 'teacher' or instructor for the course is really good and his slide decks are clear and comprehensive. One of the reasons why I managed to complete this specialisation and all the quizzes and tasks in-between in a shorter time than recommended is because I could breeze through some videos (at 1.5x speed) and refer to the slides provided after (most other programmes I have done do not offer the slides for reference). I must offer a caveat here that some of the content shared in the abovementioned courses/specialisation go beyond what is required by a teacher in a primary/secondary (or even pre-university) school but the knowledge could go a long way either for direct sharing with students who are interested or for yourselves as the teaching service evolves and transforms with the applications of AI and Big Data in learning management systems and schools.

The considerations for Algorithm Selection

What is required to build an ML model?

Cross-Industry Standard Process for Data Mining (CRISP-DM) process starting with Business Understanding

Venn Diagram of AI or ML application opportunities

Using a matrix for decision-making is something I am familiar with at work

I do have some understanding of the implications of false negatives and I can imagine it could be quite detrimental if used in the COVID context

I did multiple linear regression on Excel and tried classification for stressed faces here

Difference between AI and AGI

The 2nd and 3rd courses were much easier (less technical)

Managing ML Projects and Human Factors in AI respectively

The courses also looked at Heuristics as well as design principles when thinking about using AI 

There was also a list of biases to note

Important to apply Design Thinking (and be ready to fail fast too)

The course also looked at Data Privacy and the importance of clean and reliable data.

I have always had trouble differentiating the 2 fields AI and Data Science

Helpful graphic to see the varying involvements over the project life cycle

At work, I am already a product manager (idea owner/lead) for one of the tech prototypes that will be 'delivered' or shared with primary schools for use to implement in-class breaks but I have not had the chance to apply AI in the product just yet (although I did try to explore the use of AI for the identification or triggering of the 'service'). Next year, I will be leading the project management of another prototype (also non-AI related) that will support secondary schools in their implementation of Student-initiated Learning as part of Blended Learning and regular Home-based Learning Days. Perhaps someday in the future, I can apply some of what I have learned to support the management of AI products within MOE. However, I have to be honest, not everything is crystal clear to me as there are some areas or concepts which I probably need to relook or have some experience dealing with before having a better understanding of. Some of these more complex screenshots will be included below so in case any of my readers can enlighten me in some of the following that would be very much appreciated.

Next up on my agenda is to consider the following PD opportunities for myself in 2022: 

https://learn.aisingapore.org/ai-for-students-ai4s-educator/

https://learn.aisingapore.org/courses/learn-rpa-with-tagui-beginners-course/

I am also hoping to go for a postgraduate (Masters) looking deeper into Educational Technology as well as its assessment and evaluation at different levels. I have been eyeing several courses i.e. 

1. MA Education and Technology at UCL
2. MSc Education (Digital and Social Change) at the University of Oxford (used to be a Learning and Technology programme)
3. Master of Science (Science of Learning) at NIE (especially with its links to neuroscience at LKC School of Medicine, NTU
4. Mini Masters in Mind, Brain and Education at NTU

Clearly, I will not be able to afford some of the overseas programmes even if I got accepted there (I got offered one of the UK programmes already) so I am crossing my fingers that my scholarship application pulls through. Wish me luck!