ML Studio

Insights

Long-form articles on the craft of modern data work — machine learning models, data visualisation and statistical analysis. Practical, opinionated, and free to read.

Machine Learning
Choosing the right machine learning model: a practitioner's guide

From linear baselines to gradient boosting and deep neural networks — how to pick, train and trust the model that actually fits your problem.

14 January 2026 · 8 min read
Data Visualisation
Data visualisation that actually communicates

Charts are arguments, not decoration. A short guide to picking the right visual, removing noise and making numbers persuasive without distorting them.

3 February 2026 · 6 min read
Statistical Analysis
Statistical analysis without self-deception

Hypothesis tests, effect sizes, confidence intervals and the assumptions everyone forgets. A practical refresher for analysts who want defensible conclusions.

24 February 2026 · 9 min read
Machine Learning
Feature engineering: the part of ML that still beats AutoML

Algorithms get most of the credit, but features decide the ceiling. A field guide to encoding, scaling, interactions, leakage and the features AutoML still misses.

17 March 2026 · 7 min read
Data Visualisation
Designing dashboards people actually use

Most dashboards die from neglect within three months. A guide to building dashboards that drive decisions — hierarchy, defaults, alerts and the one chart you really need.

30 March 2026 · 6 min read
Statistical Analysis
A working analyst's introduction to Bayesian thinking

Priors, posteriors, credible intervals and why probabilistic answers communicate better than p-values in most business conversations.

18 April 2026 · 8 min read
Machine Learning
Interpreting black-box models without losing your mind

SHAP, LIME, permutation importance and partial dependence — a practical guide to extracting trustworthy explanations from complex models.

5 May 2026 · 7 min read
Machine Learning
Ensemble strategies: when one model is not enough

Stacking, blending, boosting and bagging — how to combine models so the ensemble outperforms every member, and when it is worth the complexity.

22 May 2026 · 6 min read
Data Visualisation
Data storytelling: turning charts into narratives

The best visualisations do not just show data — they tell a story. A framework for structuring analytical narratives that persuade and endure.

1 June 2026 · 6 min read
Data Visualisation
Motion in data visualisation: when to animate and when to stop

Animation can guide attention, reveal change over time, or waste it. Principles for using motion responsibly in charts, dashboards and presentations.

12 June 2026 · 5 min read
Statistical Analysis
Causal inference from observational data: a practitioner's primer

When randomised experiments are impossible, how do you still estimate causal effects? A tour of difference-in-differences, regression discontinuity and instrumental variables.

25 June 2026 · 9 min read
Statistical Analysis
Survival analysis: modelling time-to-event data correctly

When the event has not happened yet for everyone, ordinary regression fails. A guide to Kaplan-Meier curves, Cox models and handling censoring with confidence.

8 July 2026 · 7 min read
Data Visualisation
Layouts of the dashboard: choosing the right grid for the right story

A tour of the eleven dashboard layouts inside ML Studio — what each one is optimised for, where it fails, and how to pick the grid that matches your audience.

15 June 2026 · 5 min read
Machine Learning
MLOps for small teams: shipping models without a platform team

Version control, reproducible training, model registries and monitoring — the minimum viable MLOps stack for teams of one to five.

1 July 2026 · 7 min read
Data Visualisation
Accessible data visualisation: charts that everyone can read

Colour-blind palettes, screen-reader tables, keyboard navigation and the small design choices that widen the audience for your analysis.

14 July 2026 · 5 min read
Statistical Analysis
Time series analysis for people who fit regressions to everything

Autocorrelation, stationarity, ARIMA vs Prophet vs state-space models — how to forecast without insulting the statistics.

25 July 2026 · 8 min read
Machine Learning
Evaluating LLM applications: beyond vibes and screenshots

Golden datasets, LLM-as-judge, red-teaming and cost-quality trade-offs — a disciplined evaluation loop for language-model features.

4 August 2026 · 7 min read
Data Visualisation
Mapping data on maps without lying about geography

Choropleths, dot-density, hex bins and cartograms — how to choose a map that reveals rather than distorts spatial patterns.

16 August 2026 · 6 min read
Statistical Analysis
Power analysis: how many samples do you actually need?

Effect sizes, minimum detectable effects and the calculations that separate a well-designed experiment from an expensive false-negative.

28 August 2026 · 6 min read
Machine Learning
Conformal prediction: give your forecasts an honest uncertainty range

Turn point predictions into calibrated intervals, understand what coverage guarantees actually mean, and recognise when changing data invalidates them.

14 October 2026 · 4 min read
Machine Learning
Rare-event classification: why 99% accuracy can be useless

Evaluate fraud, failure and other rare outcomes using precision, recall, realistic validation and thresholds that reflect the cost of mistakes.

14 October 2026 · 4 min read
Data Visualisation
Showing uncertainty without overwhelming your audience

Use interval plots, forecast bands and thoughtful labels to communicate what the data supports—and what remains uncertain.

14 October 2026 · 4 min read
Data Visualisation
Fast dashboards for large datasets: aggregate first, draw second

Keep analytical dashboards responsive using meaningful aggregation, honest sampling and chart choices that preserve the important patterns.

14 October 2026 · 4 min read
Statistical Analysis
Missing data: when filling the gaps changes the answer

Understand MCAR, MAR and MNAR, choose defensible missing-data treatments, and check how sensitive your conclusions are to the assumptions.

14 October 2026 · 4 min read
Statistical Analysis
Multiple testing: stop a dashboard of p-values becoming a discovery machine

Distinguish exploratory screening from confirmatory testing, choose family-wise or false-discovery control, and report findings without cherry-picking.

14 October 2026 · 4 min read
Machine Learning
Cross-validation pitfalls that quietly inflate your scores

Spot the leakage, duplication and scaling mistakes that make validation scores look better than reality — and build a split you can actually trust.

18 October 2026 · 4 min read
Machine Learning
Hyperparameter tuning without wasting your budget

A disciplined search beats a thousand random fits: choose the parameters that matter, tune inside cross-validation, and stop before you overfit the validation set.

18 October 2026 · 4 min read
Data Visualisation
Colour in data visualisation: choices that carry meaning

Use colour as information rather than decoration — ordered scales for ordered data, diverging scales for deviations, and accessible palettes that survive printing.

18 October 2026 · 4 min read
Statistical Analysis
Confidence intervals: what they do and don't mean

Read intervals as ranges of plausible effects, understand what the 95% actually promises, and report uncertainty in a way decision-makers can use.

18 October 2026 · 4 min read
Machine Learning
Model drift: when yesterday's model meets today's data

Detect when a deployed model stops matching the world it serves, distinguish data drift from concept drift, and build a monitoring routine that catches both.

18 October 2026 · 4 min read
Machine Learning
Label quality: the accuracy ceiling nobody budgets for

Every model is capped by the quality of its training labels. Measure agreement between annotators, write the guideline before the first batch, and treat disagreements as data rather than noise.

19 October 2026 · 4 min read
Data Visualisation
Small multiples: many series, no spaghetti

When a single chart with twenty lines becomes unreadable, split it into a grid of identical mini-charts. Shared axes, one highlight at a time, and a fixed order turn clutter into comparison.

19 October 2026 · 4 min read
Statistical Analysis
Peeking at the experiment: why stopping early inflates your results

Checking a running A/B test and stopping the moment it looks significant quietly multiplies false positives. Decide the sample size first, or use a stopping rule built for repeated looks.

19 October 2026 · 4 min read