A collection of blog posts about machine learning
Can an AI guess which data matters without ever seeing the data? A small experiment pits a language model against two standard methods on 1994 census data, with the full notebook to run yourself.
CRISP-DM is the closest thing data analytics has to a universal playbook. Here's what the six phases actually mean in practice — and why I keep coming back to it.
Scikit-Learn is the workhorse of applied machine learning in Python. Here's what it does, why it's worth learning deeply, and a practical example to get you started.
Both entropy and Gini impurity measure how mixed a node is in a decision tree — but they differ in computation, behavior, and when each makes sense to use.
Grid Search exhaustively evaluates every combination in a parameter grid using cross-validation. Here's how it works, when to use it, and a complete implementation with Scikit-Learn.
When your parameter space is too large for Grid Search, Randomized Search samples combinations efficiently. Here's how it works and when it outperforms the exhaustive alternative.
Analytics engineer turning fragmented tidbits into confidence: messy data structured, patterns caught early, answers made plain with AI.
Copyright 2026 Christopher A. Rotunno. All Rights Reserved
