CHAPTER 24 / 40
Statistics & Optimization Foundations
3 lessons · Explore, experiment and understand.
DOJO PETYour little coding companion
DOJO / YOUR GUIDE
Let’s figure it out.
Ask me about this course or where to start.
Chapter lessons
3 lessons
Why this matters in ML
Confidence intervals, Jacobians/Hessians, and KKT duality close the foundation: every accuracy claim gets a CI, every optimizer sees curvature, and every constrained loss has a clean dual.
- bootstrap CIs
- L-BFGS
- natural gradient
- SVM dual
- Step 101~10m13
Hypothesis Testing and Confidence Intervals
Hypothesis tests and confidence intervals turn finite samples into rigorous statements about populations — the same machinery that powers A/…
Open lesson - Step 102~15m3D · 113★Most cinematic
Jacobians, Hessians, and Newton's Method
Beyond the gradient: the Jacobian is the gradient of a vector-valued function, the Hessian is the gradient of a gradient, and Newton's metho…
Open lesson - Step 103~15m3D · 113★Most cinematic
Convex Duality and KKT Conditions
Every constrained optimization problem has a *dual* — a second optimization problem whose optimum equals the original's under convexity. KKT…
Open lesson