Calculus I · 2B · exploration

Convexity, Tangent Inequalities, and Global Minima

Explore convexity, tangent inequalities, and global minima in an optional advanced note connecting Calculus I to later analysis and mathematical theory.

Course progressSupporting resourceReturn to the unit map →

Section overview

Optional advanced explorations

What this section is building

Explore convexity, tangent inequalities, and global minima in an optional advanced note connecting Calculus I to later analysis and mathematical theory.

Notice

Advanced results connect derivative evidence to error amplification, convergence speed, or global optimality under explicit hypotheses.

Decide

State the hypotheses before the conclusion and test the result on a concrete numerical or graphical example.

Avoid

Quoting elasticity, quadratic convergence, or convexity without checking units, root simplicity, or the relevant domain.

When Local Information Becomes Global

A differentiable function is convex on an interval when its graph lies below every chord, or equivalently when every tangent line lies below the graph:

f(y)\ge f(x)+f'(x)(y-x).

If $f"\ge0$ , the derivative is increasing, which leads to this tangent inequality.

**Visual study**
A convex graph lies above each tangent line. When the tangent slope is zero, that horizontal line supports the entire graph from below, so no other point can have a smaller function value.

Convexity transforms optimization. Any point with $f'(x)=0$ is a global minimum on the interval, not merely a local candidate. If $f">0$ , the minimum is unique. In higher dimensions, convex objectives and convex feasible sets form one of the most tractable classes of optimization problems.

Application

Why least squares is so manageable

A quadratic error function such as

E(m)=\sum_i (y_i-mx_i)^2

is convex in $m$ . Solving $E'(m)=0$ finds the global best-fitting slope, not a deceptive local minimum. This is one reason quadratic loss appears throughout statistics and machine learning.

Interactive checkadvanced-convex-01

For a differentiable convex function, what kind of minimum is any point where $f'=0$ ?

Your work stays on this device. No account or AI grader is used.

Show hint

Use the tangent inequality.

Attempt once to unlock the solution

Submit an answer first. The hint is available now.

Source & rights

Original instruction with traceable references.

BetterGrades-original composition declared by source handoff; owner provenance review required before public release

Reference textbooks remain rights-separated and are not published as application assets. Any direct adaptation requires separate identification and attribution.

When Local Information Becomes Global

A differentiable function is convex on an interval when its graph lies below every chord, or equivalently when every tangent line lies below the graph:

f(y)\ge f(x)+f'(x)(y-x).

f"\ge0

, the derivative is increasing, which leads to this tangent inequality.

Read this graph as text

For a convex function, a stationary point is automatically global. A convex graph lies above each tangent line. When the tangent slope is zero, that horizontal line supports the entire graph from below, so no other point can have a smaller function value. The horizontal green line touches the curve at the stationary point and stays below the graph everywhere shown. Convexity turns the local derivative statement f'(x *)=0 into the global conclusion f(x) f(x *) for every point in the interval.

Every relationship in for a convex function, a stationary point is automatically global is identified with written labels plus distinct solid, dashed, dotted, double, marker, or pattern cues; color is never the only carrier of meaning.

Why it matters: This visual shows why convex optimization is unusually trustworthy. In ordinary optimization, a critical point is only a candidate. Under convexity, the tangent inequality converts a zero derivative into a global guarantee.

Visual study

A convex graph lies above each tangent line. When the tangent slope is zero, that horizontal line supports the entire graph from below, so no other point can have a smaller function value.

Convexity transforms optimization. Any point with

f'(x)=0

is a global minimum on the interval, not merely a local candidate. If

f">0

, the minimum is unique. In higher dimensions, convex objectives and convex feasible sets form one of the most tractable classes of optimization problems.