Applied ML | DATA SCIENTIST
An explorer at heart, growing up and living in Singapore, Germany, the UK, and the USA have shaped the way I view the world and the experiences I seek. When not coding, modeling data or styling pixels, you can find me hillwalking, riding my vintage bicycle or discovering new places to enjoy a cuppa.
May trained as a neuroimaging scientist and studied brain functions in health and disease. Her love for the interdisciplinary process of discovery and translating findings into data stories paved a natural transition to applying data science in a broader context. Throughout her experience working with different data types e.g. brain (time-frequency), climate, geo-spatial, as well as unstructured (text) data, May finds joy in weaving her research experience in integrating high dimensional neuroimaging analysis with machine learning and statistical tools to discover insights.
Find out more about the work she has done and the things she's interested in discovering.
Here are few of the things I do with data and some of the tools and tech-stacks I have experience with.
Time spent on getting to know your data and its limitations is time well-spent for subsequent data analysis and interpretation.
Applying statistics to numerical analyses enable us to maximize our inference, appreciation and use of the findings.
Using algorithms that iteratively learn from data, deriving hidden insights without explicit directions has useful applications in different scenarios.
Complex patterns (e.g. object, speech) in large amounts of data can be learnt by harnessing computing power and training neural networks (e.g. using Keras | TensorFlow).
Studying how brain signals change in response to our interactions with the world in health and disease can help us understand how the brain functions.
Sharing what we do and discover is part of the scientific and social discourse.
Reviewing and publishing findings are relevant for the process of discovery.
Integrating spatial information (e.g. using GeoPandas | geojson.io | OpenStreeMap) helps us see where things are located and gives context to data.
Programming modules e.g. pandas | scikit-learn | nltk | matplotlib are the bread and butter of wrangling, modelling, and visualizing data.
An extensive range of statistical tools with great libaries e.g. for reading, manipulating, modeling, (interactively) visualizing, and reporting data.
A research 'scripting' staple. Render interactive audio-visual experience using psychtoolbox or perform time-frequency and source analyses of neuroimaging data with fieldtrip.
Great for back-end prototype/micro web development with Python
An accessible way to integrate data and scalable vector graphics (SVG) to create engaging and/or interactive online visualization.
A go-to for clean vector graphic designs and typography, as well as preparing scientific illustrations for publications.
Great discoveries, insights and creations do not happen in isolation.
~ Get in touch ~