# Stochastic Blog
> Thoughts, stories and ideas.
Public Ghost content for AI and LLM tooling. This file includes a bounded export of public pages first, then recent public posts.
Append `.md` to any post or page URL to get the content in Markdown (for example, `/example-post.md`).
## Pages
### About
URL: https://stochastic.blog/about/
Last updated: 2026-07-22T03:05:38.000Z
Stochastic Blog is where machine learning gets built in the open—carefully, from the ground up, with numbers you can check and notebooks you can run.

At the Stochastic Blog we write for people who want more than slide decks and vibes: engineers, analysts, researchers, and serious learners who prefer a working model to a buzzword.

## What this blog is

Each post is one continuous argument. We start from a question the last post left open, load a public dataset that fits on a laptop, and walk through the idea until the numbers force a conclusion. You get:

- Clear explanations of the ideas that actually show up in practice
- Short, purposeful code—not a dump of the notebook
- Real figures from a notebook that ran end to end
- A downloadable notebook so you can reproduce and extend the work
- Further reading with books worth buying, not a random link farm

The flagship project is a **48-post series**: machine learning from first principles through classical methods, deep learning, NLP, vision, reinforcement learning, LLMs, and the systems work that makes models useful. Datasets repeat on purpose, so you spend less time on setup and more time comparing ideas.

## Who it's for

- Practitioners who want intuition anchored to metrics, not metaphors alone
- Learners who are tired of tutorials that skip the failure modes
- Teams that want a shared vocabulary for overfitting, evaluation, and tradeoffs

If you only want headlines about the latest model release, this is not that site. If you want to understand *why* a number moved—and what to try next—you're in the right place.

## Why "Stochastic"

Real learning systems are noisy. Data is messy. Results move when the seed, the split, or the threshold changes. This blog treats that uncertainty as part of the craft: we measure it, name it, and design around it instead of pretending it away.

## Membership

New essays publish for **paid subscribers**. Non-subscribers can read a preview of each post; members get the full piece, the downloadable notebook, and the complete further-reading path.

[Subscribe to Stochastic Blog](https://stochastic.blog/#/portal/signup) — get the next post when it ships, and build the series with us from the first idea to the frontier.
## Posts
### Tagging and parsing English sentences
URL: https://stochastic.blog/tagging-and-parsing-english-sentences/
Last updated: 2026-10-17T03:10:42.000Z
_This post is for paying subscribers only._
### From bag-of-words to word vectors
URL: https://stochastic.blog/from-bag-of-words-to-word-vectors/
Last updated: 2026-10-12T14:10:58.000Z
_This post is for paying subscribers only._
### Augmentation, transfer, and smaller models
URL: https://stochastic.blog/post-24-augmentation-transfer-and-smaller-models/
Last updated: 2026-10-10T01:57:41.000Z
_This post is for paying subscribers only._
### Deep learning mechanics on MNIST
URL: https://stochastic.blog/deep-learning-mechanics-on-mnist/
Last updated: 2026-10-05T12:53:24.000Z
_This post is for paying subscribers only._
### Hidden states, filters, and Gaussian Processes
URL: https://stochastic.blog/hidden-states-filters-and-gaussian-processes/
Last updated: 2026-10-02T13:15:28.000Z
_This post is for paying subscribers only._
### Bayesian modeling with MCMC
URL: https://stochastic.blog/post-21-bayesian-modeling-with-mcmc/
Last updated: 2026-09-28T11:19:38.000Z
_This post is for paying subscribers only._
### Causal questions and the jobs program data
URL: https://stochastic.blog/causal-questions-and-the-jobs-program-data/
Last updated: 2026-09-27T09:23:39.000Z
_This post is for paying subscribers only._
### Forecasting household power use
URL: https://stochastic.blog/forecasting-household-power-use/
Last updated: 2026-09-21T11:15:27.000Z
_This post is for paying subscribers only._
### How models train, from gradient descent to Adam
URL: https://stochastic.blog/how-models-train-from-gradient-descent-to-adam/
Last updated: 2026-09-18T11:04:58.000Z
_This post is for paying subscribers only._
### Market baskets and association rules
URL: https://stochastic.blog/market-baskets-and-association-rules/
Last updated: 2026-09-14T11:08:04.000Z
_This post is for paying subscribers only._
### Seeing high-dimensional data in two dimensions
URL: https://stochastic.blog/seeing-high-dimensional-data-in-two-dimensions/
Last updated: 2026-09-11T11:46:58.000Z
_This post is for paying subscribers only._
### Clustering and Anomaly Detection on Handwritten Digits
URL: https://stochastic.blog/clustering-and-anomaly-detection-on-handwritten-digits/
Last updated: 2026-09-07T11:13:42.000Z
_This post is for paying subscribers only._
### Kernels, margins, and ensembles
URL: https://stochastic.blog/kernels-margins-and-ensembles/
Last updated: 2026-09-07T08:15:01.000Z
_This post is for paying subscribers only._
### Linear models from least squares to quantiles
URL: https://stochastic.blog/linear-models-from-least-squares-to-quantiles/
Last updated: 2026-09-04T11:12:26.000Z
_This post is for paying subscribers only._
### Validation, imbalance, and tuning done right
URL: https://stochastic.blog/validation-imbalance-and-tuning-done-right/
Last updated: 2026-08-31T11:10:29.000Z
_This post is for paying subscribers only._
### First classifiers and how to score them
URL: https://stochastic.blog/first-classifiers-and-how-to-score-them/
Last updated: 2026-08-28T11:15:21.000Z
_This post is for paying subscribers only._
### Cleaning and shaping the Airbnb listings
URL: https://stochastic.blog/cleaning-and-shaping-the-airbnb-listings/
Last updated: 2026-08-24T11:28:40.000Z
_This post is for paying subscribers only._
### Hard problems and heuristic answers
URL: https://stochastic.blog/hard-problems-and-heuristic-answers/
Last updated: 2026-08-20T12:00:28.000Z
_This post is for paying subscribers only._
### Graph Algorithms on the World Flight Network
URL: https://stochastic.blog/graph-algorithms-on-the-world-flight-network/
Last updated: 2026-08-14T12:26:34.000Z
_This post is for paying subscribers only._
### Subproblems, strings, and search tricks
URL: https://stochastic.blog/subproblems-strings-and-search-tricks/
Last updated: 2026-08-10T12:19:51.000Z
_This post is for paying subscribers only._
### Sorting, hashing, and sketches on 370,103 words
URL: https://stochastic.blog/sorting-hashing-and-sketches-on-370-103-words/
Last updated: 2026-08-07T12:24:24.000Z
_This post is for paying subscribers only._
### Information Theory in One Sitting
URL: https://stochastic.blog/information-theory-in-one-sitting/
Last updated: 2026-08-20T12:36:02.000Z
_This post is for paying subscribers only._
### Statistical tests without the traps
URL: https://stochastic.blog/statistical-tests-without-the-traps/
Last updated: 2026-07-31T12:18:41.000Z
_This post is for paying subscribers only._
### Probability and estimation on census data
URL: https://stochastic.blog/probability-and-estimation-on-census-data/
Last updated: 2026-07-27T13:14:31.000Z
_This post is for paying subscribers only._
### The linear algebra and calculus behind every model
URL: https://stochastic.blog/the-linear-algebra-and-calculus-behind-every-model/
Last updated: 2026-07-24T15:06:15.000Z
_This post is for paying subscribers only._
### The first six ideas in machine learning
URL: https://stochastic.blog/the-first-six-ideas-in-machine-learning/
Last updated: 2026-07-24T15:06:40.000Z
_This post is for paying subscribers only._