Q – What does your day job entail?
First, I open MATLAB and any other IDEs needed that day since my work life revolves around coding! I find ways to share that code with the community (in the form of demos, videos, at conferences like Big Data London!) and I share the code with MATLAB developers in the data science and deep learning / machine learning areas to help them understand the user experiences and improve the software.
Q – When did you realise that you were suited to a career in data?
During my PhD research, I used machine learning to classify microscope images based on fluid concentrations (this could lead to identifying diseases from droplets of bio fluids). It was super exciting, and I fell in love with machine learning, data science, and coding. I also didn’t have a lot of funding, so I traded other PhD students for the fun parts: they did my Scanning Electron Microscopy, I did their data analysis and programming!
Q – What piece of advice would you give to someone starting out in data?
Try everything! Get your hands dirty and try as many languages, packages, frameworks as possible (try not to get stuck with one go-to model or language, keep an open mind). Try Kaggle competitions or challenges (but for the experience of building models on different types of data sets and learning from your peers… not the glory of winning!) Talk to as many people as you can and engage in communities (meet ups, social media, forums, etc) - you learn so much from others!
Q – What do you think will be the hot topics at Big Data LDN 2019?
I expect to hear loads about AI this year, especially on model explainability and interpretability, streaming and online model updating, and use of AI in new applications.
Q – What can people expect from your session?
The theme for my session is all in the title “would you trust your model with your life?” Many of us have experience building models, but what happens when you need to take that next step and put the model into practice? There’s even more to think about when you are building models for safety-critical applications like automated driving and smart medical devices. We’ll discuss these challenges and techniques to address them like model explainability, testing, simulations, and more.