7: Optimization
( \newcommand{\kernel}{\mathrm{null}\,}\)
- 7.1: Introduction to Optimization
- What is optimization? Introduction to cost and parameter space.
- 7.2: Single-Dimension Continuous Optimization
- Optimization of known and continuous single-variable functions, using the zero-derivative test and Newton's method.
- 7.3: Multi-Dimensional Continuous Optimization
- Methods of finding the optimization of a multidimensional cost function that depends on more than one variable: steepest-descent, conjugate gradient, and Newton's second-order method.
- 7.4: Linear Programming
- Optimization in cases where cost is a linear function of multiple parameters, comprising inequality and/or equality constraints.
- 7.5: Integer Linear Programming
- The branch-and-bound method for solving optimization problems that involve continuous cost and constraint functions but allow only integer solutions.
- 7.6: Min-Max Optimization for Discrete Choices
- The min-max method: finding the smallest normalized deviation from the peak performance across objectives, as a way to select the optimal solution in a situation which has multiple discrete candidates.
- 7.7: Dynamic Programming
- Introduction to dynammic programming approach of solving optimization problems with a high number of unknowns, specifically focusing on shortest-path problems.
- 7.8: Solving Dynamic Programming on a Computer
- Using the value iteration algorithm to efficiently solve shortest-path optimization problems with a computer.