Take 10 with... Yun Sing Koh

Associate Professor Yun Sing Koh from the School of Computer Science gives us 10 minutes of her time to discuss her research into mining data streams.

1. Describe your research topic to us in 10 words or less.

Developing adaptive predictive machine learning algorithms for data streams.

2. Now describe it in everyday terms!

Many applications require mining of data streams where the data comes from various sources, such as weather sensors, computer network traffic, and web searches. This vast amount of data provides an excellent opportunity to discover rich knowledge automatically. Machine learning uses data to find patterns and to predict or infer new knowledge. Mining of data streams enables future states of a stream to be predicted by learning models based on historical data and then adapting such models to changes that take place in the stream over time. My research aims to develop these adaptive machine learning techniques, which can be applied to different application domains.

3. What are some of the day-to-day research activities you carry out?

There is wide range of activities, including investigating current research, brainstorming, designing prototypes, coding and experimentation, validation and testing, discussing research with university and industry collaborators, and supervising postgraduate students.

4. What do you enjoy most about your research?

The different things I am learning on a day-to-day basis are fascinating. Brainstorming with collaborators and PhD students is invigorating. I particularly enjoy supervising and mentoring students, being able to see and help them grow their research is wonderful.

5. Tell us something that has surprised or amused you in the course of your research.

I am always amazed when people I meet on a day-to-day basis outside academia, who want hear about my research and the keen interest and questions they have on my research and the implications it will have on their daily life.

6. How have you approached any challenges you’ve faced in your research?

Some of the approaches I try include looking at a problem or a challenge from different perspectives. For example, when we discuss machine learning research there are always questions on biases and ethical implication.  These are questions we do ask of our techniques, but the way it is presented needs to be different for different applications.

7. What questions have emerged as a result?

How do I understand the model itself and what is it actually modelling?

8. What kind of impact do you hope your research will have?

I am trying to advance machine learning algorithms. With new machine learning techniques, we can produce far better results than we could previously across various domains and industries.

9. If you collaborate across the faculty or University, or even outside the University, who do you work with and how does it benefit your research?

I work with various industries from environment, software, and health. One cool thing about machine learning is its broad applicability. It can be applied across so many domains to help in a wide range of problems. I also collaborate with a number of my colleagues within the school, New Zealand and international researchers. The ability to brainstorm and discuss research is crucial to help grow innovation and understanding of the area. 

10. What one piece of advice would you give your younger, less experienced research self?

As a female computer science researcher, it is great to have mentors that have paved the way. I was very lucky to have fantastic mentors such as Professor Gill Dobbie and Dr Pat Riddle.  I believe this has made a big difference in my career.