Intuition · Theory · Proof

Intuition Comes First.

Statistics and machine learning are the same inquiry. Different eras, different names.
"The statistical community has been committed to the almost exclusive use of data models. This commitment has led to irrelevant theory."

— Leo Breiman, 2001

Read: The Two Cultures

What This Place Is

A laboratory where algorithms are disassembled, not just applied. A library where foundational arguments are made interactive. A set of case studies where mathematical models meet real data. Not a course. Not a tutorial. A point of view, made navigable.

§ I

The Laboratory

Strip away the black-box libraries. Explore the numerical engines underneath — Babylonian roots, gradient descent, Newton-Raphson — as living, interactive objects.

Enter the Lab
Lab 06
Gradient Descent
Optimizing...
Click to drop point
Metrics
Loss: ∇J(θ)
2470.623
Iteration (Epoch)
0 / 150
Fig. 1 — The Laboratory
§ II

Case Studies

Real datasets. Decisions made visible. Watch a model move from raw numbers to structural insight — step by step, assumption by assumption.

Browse the Archive
Case 04
Time Series Analysis
Stochastic Forecast
ARIMA Parameters
Damping (ρ)0.85
Seasonality15
Uncertainty1.5
Historical (t≤0)Forecast (t>0)
ε2MSE
26.76
ε2RMSE
5.17
|ε|MAE
4.70
𝒜AIC
97.9
Fig. 2 — Case Studies
§ III

The Library

Foundational papers in statistics and machine learning. Read the original arguments that shaped the field, annotated and contextualized.

Enter the Reading Room
Volume I
The Reading Room
Curated Texts
Ref. 2001

Statistical Modeling:
The Two Cultures

Leo Breiman
Ref. 1948

A Mathematical Theory of Communication

Claude E. Shannon
Fig. 3 — The Library

Mathematics in motion.

The Null Hypothesis treats algorithms as engines to be understood, not tools to be consumed. Every concept here has a visual proof. Every model has a mechanism you can see.

Read the Institutions