Sarada Lee, Co-founder and Applied AI Researcher at Perth Machine Learning Group (PMLG).
What is your academic background?
I studied a Master in Professional Accounting at The HongKong Polytechnic University (PolyU). Other positions/post nominals include: Visiting Scholar (University of San Francisco, Data Institute); Conjoint Fellow (University of Newcastle, Australia, School of Medicine and Public Health); GAICD (Australian Institute of Company Directors); FCCA (Association of Chartered Certified Accountant); and HKICPA (Hong Kong Institute of Certified Public Accountant).
Have you completed any other training in data science?
I have completed ai / Deep Learning – Part I & II, University of San Francisco (USF), Data Institute; lots of free online resources (ie videos, forums, blogs, papers) and have also learnt a lot via Meetup.
What advice would you give others looking to pursue a data science career?
- Don’t listen to those (including yourself) who say you can’t do it
- Surround yourself with people with similar interests and who are smarter than you (This is how PMLG was founded)
- Learn from the best teacher/mentor you can find with the teaching style that most suits your way of learning
- It is ok if you can’t understand everything and you don’t have to. It’s more important that you understand how you can learn your new subject.
What sparked your interest in working with data?
I love looking for information in data. When you have good information, you can support good decision-making.
How did you come to work in your current role?
Totally unplanned. I was made redundant as an accountant in late 2016 and was drawn to the field as a natural extension of my previous analytical role.
What sort of projects have you been working on?
From a technical aspect, here are some highlighted projects:
• Computer Vision (Segmentation): Solar panel and items which affect electricity consumptions; Immune cells analysis
• Computer Vision (Classification): Plant seedings classification; Superheros’ merchant classification
• Tabular Data (Regression): Predictive maintenance for crusher’s mantle wears; Propensity to fund mortgages
• Natural language processing (Generative): Created poetry (in English and Chinese) which were exhibited in Fringe World 2019, BitLit: Machine Poetry Corner**
• Generative Adversarial Networks: Created interactive art to display in Fringe World 2020, GANify: Funhouse Mirror Machine** which is received the visual art award; Enhanced low resolution images to high resolution images
What tools/platforms do you use in your work?
• GPU Compute: AWS, Goolge Colab or local GPU compute servers
• Machine learning: Python, fastai, PyTorch, Jupyter Notebook, SkiLearn, Pandas
• Other tools: Github, Ubuntu, Numpy, Matplotlib, spreadsheet, annotation tools, R, SQL
What has been a highlight of your data science career so far?
Being a Women in Technology WA Tech [+] 20 Award Winner in 2019.
What challenges have you faced as a data scientist?
• Some people still think you need to have a PhD to do it. They ignore other important aspects, such as communication & project management skills, compliance and the ability to deliver.
• Decision makers are very conservative in adapting machine learning or artificial intelligent into their businesses. I address some of their concerns in this article.
• The barriers to entry into this field as well as barriers to exit from a professional career.
What are some of the big areas of opportunity/questions you want to tackle in this space?
Using AI to speed up medical diagnosis and offer more tailored treatment for patients. Currently, I am working on computer vision for pathology in collaboration with Telethon Kids Institute, University of Newcastle Australia, University of San Francisco Data Institute and Wicklow AI Medical Research Initiatives.
What excites you most about recent developments in Data Science?
The release of fastai v2 library and its book called “Deep Learning for Coders with fastai and PyTorch: AI Applications without a PhD” by Jeremy Howard and Sylvain Gugger later this year.
What does the future of data science look like?
More diverse because “YOU” can be part of it. The field is expanding so rapidly at the moment that it can be hard to track. What we should expect is faster, cheaper and more impressive results from the deep learning track and increasing automation and commoditisation in the traditional data science area.
For people considering a career in data science, what is one piece of advice you would give?
Start now! Be flexible and resilient.