Statistical learning and prediction
The statistics that machine learning is built on: the bootstrap, the lasso, random forests, and cross-validation, worked out by hand on a small dataset. What the software does, shown before you let it do it.
Articles

How a random forest makes a decision
A decision tree asks one question at a time and is easy to read and easy to fool. Grow hundreds of trees on resampled data, let each see only a random subset of the variables, and take a vote: that is a random forest, and the reason it works is the arithmetic of averaging.

What the lasso does to a regression
Add a penalty on the sum of the absolute coefficients to least squares and something unexpected happens: coefficients shrink, and then, one by one, become exactly zero. Robert Tibshirani's method of 1996, the diamond and the circle, and why genomics needed it.

The bootstrap in one page
Draw your own sample again and again, with replacement, and compute the statistic each time. The spread of the results is the standard error you could not derive. Bradley Efron's method of 1979 explained, with the reason 37 per cent of the data sit out of every resample.
In the masterclass
The Bootstrap and Cross-Validation
Training and test error, bias and variance, the bootstrap, and k-fold cross-validation
Ridge, Lasso, and Regularization
Shrinkage, ridge regression, the lasso, soft thresholding, and choosing the penalty by cross-validation
Trees, Random Forests, and Boosting
Decision trees, bagging, random forests, gradient boosting, and when to use them