Let be a function defined in an open interval containing . The derivative of a function at is denoted by and is defined by:
Differentiation Rules:
- Constant rule
- Power rule
- Sum rule
- Difference rule
- Constant multiple rule
- Product rule
- Quotient rule
Chain Rule If we have then
Derivatives of Exponentials & Logarithms
Numerical Differentiation
- One-sided (forward) difference:
- One-sided (backward) difference:
- The central difference:
Fermat’s Theorem
If has a local extremum at and is differentiable at , then .
Extreme Value Theorem: A continuous function over a closed, bounded interval has an absolute maximum and minimum. Each extremum occurs at a critical point or an endpoint.
Rolle’s Theorem
Let be a continuous function over the closed interval and differentiable over the open interval such that . There then exists at least one such that .
Mean Value Theorem
Let be continuous over the closed interval and differentiable over the open interval . Then, there exists at least one point , such that:
- If for all x$$\epsilon (a,b), then is an increasing function over
- If for all x$$\epsilon (a,b), then is an decreasing function over
Convex vs Concave Functions
| Feature | Convex | Concave |
|---|---|---|
| Visual Shape | Cup / Smile () | Cap / Frown () |
| Midpoint Test | Line is above the curve | Line is below the curve |
| Optimization | Easy to find the minimum | Easy to find the maximum |
| Second Derivative |