Interview with Dr Eriita Jones
This blog post is the first of the AICN Local Interview series and is featured in Issue 1 of the #AICollaborative Network Newsletter. View a full copy of the Newsletter. Forward the Newsletter to a friend. Subscribe to receive the Network Newsletter in your inbox.
What follows is the transcript of an interview with Dr Eriita Jones conducted by Yolanda from AICN. Eriita is a Research Fellow at the School of Information Technology and Mathematical Sciences, University of South Australia.
Eriita’s research interests involve the search for liquid water and habitable environments on Mars. She uses multivariate statistical analyses of remote sensing data and thermal modelling.
We talked about the importance of collaboration, mentorship, Star Trek, and her journey into the world of machine learning.
AICN: Hi Eriita! Welcome, and thank you so much for meeting with me and having a chat :-)
Eriita: No worries, thank you for having me!
AICN: To begin with, could you tell us a little bit about yourself? Your PhD from ANU is in astrophysics, can you tell us about what led you to study that field?
Eriita: I have had a passion for space ever since I was little. I spent my first few years in the Pilbara, in northern Western Australia. The stars there were amazing, the night sky was beautiful, so I’ve always had that passion for space. Star Trek was also a huge inspiration for me growing up (and still is) and the whole search for life, and search for extraterrestrial life (SETI) is really exciting to me.
AICN: Me too! I grew up in Adelaide and have always been drawn to the stars, but I’m more of a Star Wars girl myself…! So after your postdoc at the University of Western Ontario in Canada, you returned to Adelaide. How did you find it over there and what led you back to Australia?
Eriita: I loved doing my postdoc in Canada. I was part of the Centre for Planetary Science and Exploration. It was really exciting to be part of a large funded group like that because it didn't seem that there was anything like that in Australia at that time.
After my postdoc finished, I was deciding where to go next. Of course there are many more opportunities for my kind of space research in North America, but I was so lucky to have had such a great education in Australia, and I wanted to contribute and give something back. I have family here in Australia and one of my former mentors was in Adelaide - so I decided to head back here.
But also, both Adelaide and Canberra are the space capitals of Australia. There are so many entrepreneurs, space startups and space businesses in Adelaide (not to mention Australia's only launch site nearby), so it’s a really exciting place to be for space!
AICN: You mentioned having a mentor. What are your thoughts on mentorship? Have you had strong ones throughout your career and how important are mentors for professional and personal development?
Eriita: My strongest mentor throughout my whole life would be my mum. She’s a great inspiration for me because she’s so resilient and has so much wisdom. She has been my main mentor.
Within my academic field, I’ve had mentors for different parts of my journey. I think mentorship has a huge role particularly as a woman in STEM, it’s so important to see other women treading that path in front of you, to see that women can do anything despite adversity, and to see examples of women rising to the top and doing amazing things in all areas of STEM. I think that’s really important
AICN: So is imposter syndrome something you have experienced, or see in other people and do you have strategies to counter that feeling?
Eriita: It’s definitely something I have experienced personally, and my friends and colleagues and I have all shared conversations about this issue from time to time.
'Imposter syndrome' is really challenging to address, as the representation of women in STEM fields is typically very low. The higher you go up the different career progression levels, the fewer and fewer women there are. For example in the academic career track, you get a lot of women drop off just before PhD, a lot of women dropping off during PhD for all sorts of reasons, and after PhD as well, so as you progress the proportion of women becomes smaller and smaller. I observed this first hand with my female colleagues in physics. This definitely leads to imposter syndrome being fostered.
To combat it personally, self-belief is hugely important. There is that role of seeing female role models, but being able to believe in yourself and having that internal drive for “nevertheless, she persists”. You keep plodding on, being yourself and doing your best.
But really to address this issue requires that it is not seen as a ‘female problem’ but really as a societal problem that we all have the responsibility to improve. The loss of women from STEM careers and the many biases that women within STEM careers face is a major issue, and we are losing or excluding a major source of talent.
AICN: Do you have role models in STEM or otherwise that you look up to?
Eriita: Does it have to be a real person? Haha. In terms of real people in STEM, I guess for me every female who is contributing and sharing scientific knowledge, and doing so with wisdom and compassion, and despite the adversities they may face, is someone who I look up to. And from the world of science fiction - Captain Janeway is one of my huge role models, her character was a great female scientist! I actually met her during my PhD and I have a necklace that I wear when I’m doing something tricky and I just want to believe in myself.
AICN: You are a true Trekkie! So you are presenting at the AI Collaborative event next month. Can you give us the gist of your current research program, why you think it’s relevant, and why people who are not familiar with your field should sit up and notice?
Most broadly, my interest is in using satellite data to search for hidden characteristics of the surface and subsurface that we can’t see at visual wavelengths. This could be searching for subsurface liquid water on Mars, environments that can support life, or finding the resources needed to sustain a human presence on Mars or the Moon, or searching for groundwater in arid environments in Australia (to support agriculture, remote communities, etc.). Often what that means in practice is that you are working with large multivariate, complex data sets.
My interest in AI and machine learning is about how it enhances our ability to understand these big complex data sets where you have many hidden associations between the different variables. Machine learning has a role to play in helping us tease out those relationships.
AICN: How long have you been working with machine learning and how did you come across it in the first place?
I first heard about neural networks when I was doing coursework as a PhD student, and we were learning about land-cover classification using satellite imagery. We came across a paper where someone had used neural networks to do clustering and unsupervised classification of data. The results were amazing. The accuracy was much higher than what we were producing and I thought, “Wow, how does that work?“. I tried to read some papers but I didn’t really understand how the algorithms worked.
I didn’t come back to it until last year, when I met Mark McDonnell and some others, and started learning from them about machine learning. I started to really see the power and utility of these techniques, and started understanding more about how the algorithms work.
AICN: It sounds like collaboration has been important for your learning and your work!
Eriita: Yes, collaborations are huge. Science in particular needs and benefits from collaborative work. You need peers to assess what you do and test the validity of your research (peer review), and also to have conversations and inject new ideas. Particularly when we have such a wealth of data that are so complex, you often don’t have all the skills you might need to answer the questions yourself.
When we’re looking at really big questions, holistic skill sets are the best way to approach it. So, if someone else has expertise that you don’t have, you learn from them, then you share your new knowledge with other colleagues and it ripples out. It’s for everyone’s benefit!
AICN: What other skill sets have you found useful for your work?
Eriita: The main area at the moment would be remote sensing and data analysis and statistics, for analysis of big spatial and spectral data sets. There are always new analysis techniques, so in recent collaborations I’ve been learning from data scientists who have exposure to techniques that I hadn’t used before.
Personally, I have also been learning a lot more about doing things programmatically. There are some key software packages that people in my field use. I’ve been learning about the benefit of stepping away from that more and using open source code, developing your own tools, sharing them openly. I love open source publication, so continuing going down that avenue in other areas has been a great learning experience for me.
AICN: Learning something outside of your domain can be difficult. How did you find support in your learning? Especially with machine learning, it’s such a fast-changing field, how do you keep up with the latest developments?
Eriita: That comes back in part to the mentorship aspect. You might be really lucky to meet people who also have those values of collaboration and mentorship that are willing to spend time with you to teach and share. Sometimes people don’t have that mentality, or they might feel like they don’t have the time because they are under a lot of pressure in their own field to keep producing. So to find people with whom you can share and collaborate is really great.
Personally, because I don’t have the machine learning expertise myself, I am always learning. I have lots of conversations with my collaborators, I’m guided by their knowledge and they direct me to resources as well to skill up and learn myself. I’m lucky to have great mentors in the ML area. It’s changing so rapidly as a field and what is at one point the best performing technique or approach quickly becomes outdated, so it’s great to have a guide on this journey.
Apart from that, there are so many resources online now. I do Coursera courses and skill up in different areas that way.
AICN: What stage is your research program now and where do you see it progressing?
Eriita: My research program is pretty broad. I am working on a number of different questions that I’m trying to answer. I’m really interested in using the machine learning aspect with planetary data. So at the moment, we are learning in what ways we can use machine learning in finding resources for exploration, life and habitability.
There is a huge role for ML to enhance the ability to do science in that area. I hope to use the techniques more broadly in the future and I’m really excited to see where it goes!
AICN: If you had the power and influence to change anything affecting your work in Adelaide, what would it be?
Eriita: I think one of the biggest issues is that there is a lot of short term contracts. The continual short term funding and contracts is challenging and can hold back progress in any field. When you continually have short contracts, and even some of that time is spent trying to obtain your next funding source, that really limits the amount of research you can do and the scale of projects you can initiate. Not to mention it's really hard on individuals and families to not have longer term financial stability.
It’s a big challenge and universities are going more down that route which is a major problem. More longer term funding and positions will be a great benefit, as it enables you to address and tackle larger research projects with more depth.
AICN: Thank you so much for chatting with me Eriita. I really look forward to seeing your presentation on Dec 6th!
Eriita: Thanks Yolanda, it’s my pleasure. It’s great to become connected with the AICN community and learn from everybody!
Dr. Eriita Jones
Eriita has been a member of the Artificial Intelligence Collaborative Network since October 2018. Find out more in Eriita's biography.