Practical AI & ML
First-principles thinking for applied machine learning.
Embedded Files
Hi, I’m Greg
Hi, I’m Greg
I’m a Data Scientist with over a decade of experience building ML systems in industry and a PhD in Applied Statistics from the University of Cape Town, South Africa. Find out more about my journey from academia to industry here.
I created this space to share my thoughts on AI, Data Science (DS), and Machine Learning. My interests are broad, but this site will host a collection of technical articles, tutorials, notes, reflections, and structured thinking on applied ML, DS, and AI. My goal is to share my thinking around these concepts, and in the process help Data Scientists, AI & ML Engineers, Product Managers, and Technical Leaders working with applied ML (and indeed, anyone interested in the field).
I'm always interested in hearing how others are applying AI/ML -- the successes, and lessons learned. Feel free to get in touch, or to subscribe to get notified when I publish a long-form article on this site.
My Philosophy
My Philosophy
I approach AI & ML like a scientist: forming hypotheses, testing assumptions, and interpreting outcomes objectively, even when data is messy or organisational processes are inconsistent.
This lens guides everything I write, from project reflections, technical implementations, to critiques of emerging trends.
Recent Activity
Completing the series, here we discuss the ways in which models themselves can account for Class Imbalance, why it works, and in what situations.
Class imbalance can significantly impact model performance, particularly when the minority class is the one we care most about. In this post, we'll explore the data sampling approaches available to help rebalance our training data.
Explore how Class Imbalance in the target can present significant challenges in ML models, and what to do about it.
Page updated
Google Sites
Report abuse