Summary of "제4장: 연쇄 법칙과 곱미분 법칙의 시각화 | 미적분학의 본질"

Main ideas and lessons

When modeling real-world quantities, functions rarely appear alone; they are commonly combined or modified using a few core operations.
To differentiate any complex expression, it’s enough to understand how derivatives behave under three basic combinations:
1. Addition of functions
2. Multiplication of functions
3. Composition (putting one function inside another)
The emphasis throughout is that derivative rules should not be memorized mechanically; they can be understood as the result of how the output changes when the input is nudged by a tiny amount (using the intuition behind differentials like (dx)).

Start from the definition/intuition of a derivative:
- Nudge the input (x) by a tiny amount (dx).
- Track how the function value changes by a corresponding tiny amount (e.g., (df)).
- Use proportional reasoning (changes are approximated via derivatives of simpler pieces).
- Finally compute (\frac{df}{dx}) as a ratio of tiny changes.

If a function is built as a sum:
- (f(x) = g(x) + h(x))
Instruction:
- Differentiate term-by-term using tiny-change logic:
  - The total tiny change in the sum equals the sum of the tiny changes.
- Therefore: [ \frac{d}{dx}[g(x)+h(x)] = g’(x) + h’(x) ]
Example reasoning used:
- (f(x)=\sin x + x^2)
- The derivative becomes:
  - (\cos x + 2x)

If a function is built as a product:
- (f(x) = g(x)\,h(x))
Instruction (visualized as an “area of a box”):
- Interpret (g(x)h(x)) as area where:
  - one side length changes like (g(x))
  - the other side length changes like (h(x))
- When (x) is nudged by (dx):
  - the width changes by about (g’(x)\,dx)
  - the height changes by about (h’(x)\,dx)
- The area change has components:
  - “bottom strip” proportional to (g(x)\,h’(x)\,dx)
  - “side strip” proportional to (h(x)\,g’(x)\,dx)
  - a corner term proportional to (dx^2), which becomes negligible as (dx \to 0)
- Divide by (dx) to get: [ \frac{d}{dx}[g(x)h(x)] = g’(x)h(x) + g(x)h’(x) ]
Example reasoning used:
- For (f(x)=\sin x \cdot x^2):
  - derivative is (\cos x \cdot x^2 + \sin x \cdot 2x)
Included mnemonic idea:
- “left d right + right d left” (conceptually matches the two terms in the product rule)

If a function is built by composition:
- (f(x)=g(h(x)))
Instruction (visualized as multiple “number lines” / stages):
- Think of (x \to h(x)\to g(h(x))).
- When (x) is nudged by (dx):
  - (h(x)) changes by (dh \approx h’(x)\,dx)
  - then (g(h)) changes by:
    - (dg \approx g’(h)\,dh)
- The derivative comes from multiplying the “outer” sensitivity by the “inner” sensitivity: [ \frac{d}{dx}[g(h(x))] = g’(h(x)) \cdot h’(x) ]
Key conceptual point highlighted:
- The chain rule reflects the real cancellation of the intermediate differential ratio when expressing output change in terms of the original (dx).
Example reasoning used:
- (f(x)=\sin(x^2))
- Outer derivative: derivative of (\sin(\cdot)) is (\cos(\cdot))
- Inner derivative: derivative of (x^2) is (2x)
- Result:
  - (\cos(x^2)\cdot 2x)

Most complicated functions can be “peeled open” layer-by-layer because they ultimately come from:
- adding
- multiplying
- composing
Even if expressions become “monstrous,” knowing these three derivative behaviors lets you systematically compute derivatives by repeatedly applying the rules.

Primary speaker: The video narrator/author (no name explicitly given in the subtitles)
Referenced/partner source: Brilliant.org (promoted in the closing segment)