Learn the multilayer perceptron by playing: tune a single neuron to solve AND and OR, discover why XOR needs hidden layers, activations and backpropagation.

Learn generalized linear models by playing: the three GLM components, matching outcomes to Normal, Binomial, Poisson and Gamma families, and a link-function lab.

Learn Gaussian process classification by playing: kernels, length scale and noise in a live GP playground, the latent function and link, and GPC code in Python.

Learn Bayesian linear regression by playing: Bayesian vs frequentist, Bayes’ theorem, and a live prior-to-posterior playground, with PyMC code.

Learn stochastic gradient descent by playing: loss functions, batch vs mini-batch vs stochastic, a live learning-rate race, and SGDClassifier code.

Learn stepwise regression by playing forward selection on real data: pick features by p-value, compare forward, backward and bidirectional methods, AIC and BIC.

Learn support vector machines by playing: hyperplanes, margins and support vectors, the C parameter on real data, the kernel trick, and SVM code in Python.

Learn MAE, MSE, RMSE, max error, R² and adjusted R² by playing: drag an outlier and watch every metric react, then compute them with scikit-learn.

Learn the confusion matrix, accuracy, precision, recall, specificity, F1, MCC, log loss and ROC AUC by playing with a live threshold and ROC curve on real data.