Trends and tricks
"The Data Science industry tends to attract folks with strong technical insights, creativity, and a driving passion for the work. This is an awesome combination, and colleagues continually motivate and share learning with one another."
As a Data Scientist at a leading South African machine learning consultancy, Tim's taken multiple deep learning projects from concept through to production, across the domains of unsupervised learning, natural language processing and sequence-to-sequence modelling. Advanced calculus, linear algebra and Bayesian statistics are also just some of his favourite things. And while that might sound overwhelming, his flair for teaching means that he's able to explain even the most complex of concepts to any audience.
Tim-Piggott, iXperience 2018 Data Science Head Teacher
I’m currently working on a project for a client that segments customer behaviour, and profiles each customer in terms of a series of action vectors over the course of their lifetime with the client. I’ve also worked in the natural language processing domain to represent large, unstructured bodies of text (e.g. crypto white papers) in terms of graphs; modelling thematic connectivities and high-level structure.
Firstly, the variety of work and applications! Machine learning/ deep learning provides an amazingly rich and flexible toolset across domains, and adapting powerful methods to new contexts is an awesome experience. Secondly, the people. The Data Science industry tends to attract folks with strong technical insights, creativity, and a driving passion for the work. This is an awesome combination, and colleagues continually motivate and share learning with one another.
I see clearer best practices emerging in Data Science, and increased regulation. Data Scientists will need to have a firm theoretical foundation, strong programming skills, and flexibility across platforms. They'll also need the ability to efficiently navigate large volumes of data, and integrate data sources, be at home with cloud computing, and feel ready to take their models into a production environment. Lastly, they will need to be flexible enough to adapt and optimize learning approaches across different contexts.
Students will leave with a firm grasp of the broad theoretical foundations underlying deep learning. They'll have a working proficiency in Python for data science and deep learning, exposure to Keras and Tensorflow, and experience in serving models for production environments. Relational databases, bash scripting and cloud computing will also be areas that we'll cover.
Launch yourself into the material, and take advantage of the awesome resources and community around deep learning and data science online. There are so many folks who really want to make this knowledge as accessible as possible. Work with your peers and don’t let yourself get stuck on a topic – there are no silly questions!
I used to be music-obsessed, played the clarinet, and sang in choirs.