Operators, product owners, and people affected by the classical baselines decision.
Classical ML · weeks 6
Linear, probabilistic, neighbor & margin models
Support tickets need explainable routing under a 15 ms latency budget.
Opening production brief
A measurable problem, before a model.
regularization, log odds, Bayes rule, feature scaling with version, owner, event time, and availability contracts.
Support tickets need explainable routing under a 15 ms latency budget. Every solution must state latency, cost, capacity, and fallback limits.
Offline evidence plus a deployable resolution: A justified baseline across regression, Naive Bayes, kNN, and SVM.
Frame and baseline
Turn the production problem into explicit data, metric, and baseline contracts.
- 01↗
Linear regression as a production baseline
A linear baseline is valuable because its assumptions, errors, and serving cost are easy to see.
65–90 min · checkpoint · deterministic lab - 02↗
Logistic regression, odds, and calibration
Logistic regression adds evidence in log-odds space; calibration and policy give that score operational meaning.
65–90 min · checkpoint · deterministic lab
Build and stress
Formalize the machinery, test counterexamples, and expose system limits.
- 03↗
Naive Bayes under sparse data
Naive Bayes turns sparse counts into additive evidence—quickly, smoothly, and with assumptions in plain sight.
65–90 min · checkpoint · deterministic lab - 04↗
kNN, distance, and local decisions
In kNN, representation defines the neighborhood, and the neighborhood defines the decision.
65–90 min · checkpoint · deterministic lab
Resolve and operate
Join model behavior to architecture, observability, rollout, and rollback.
- 05↗
SVMs, margins, and kernels
SVMs learn from the boundary’s hardest points; kernels change geometry, not scores into probabilities.
65–90 min · checkpoint · deterministic lab - HW↗
Explainable ticket-routing baseline
Compare linear/logistic regression, Naive Bayes, kNN, and SVM under one evaluation and serving contract.
Rubric · staged hints · reference resolution
Prerequisite guidance
2 concepts make this faster.
Prerequisites are guidance, never hard gates.