最优化-lec1 Introduction Preliminaries 最优化理论与方法课件（英文版－南京大学）.pptVIP

Chapter 1 Introduction Preliminaries * Optimization Optimization models attempt to express, in mathematical terms(数学术语), the goal of solving a problem in the “best” way running a business to maximize profit, minimize loss, maximize efficiency, or minimize risk designing a bridge to minimize weight or maximize strength selecting a flight plan for an aircraft to minimize time or fuel use Optimization models have been used for centuries, as their purpose is so appealing, and in recent times, they have come to be essential The concept of optimization is well rooted as a principle underlying the analysis of many complex decision or allocation problems Optimization Process * real world problem algorithm, model, solution technique computer implementation analysis numerical methods verification 验证 validation, sensitivity analysis * Optimization One approaches a complex decision problem involving selection of values for a number of interrelated variables(关联变量) focus attention on a single objective(单目标) designed to quantify performance and measure the quality of the decision(衡量决策品质) the objective is maximized or minimized subject to the constraints that may limit the selection of decision variable values(决策变量值) A particular optimization formulation should be regarded only as an approximation Learn to identify and capture the important issues of a problem * Types of Problems ☆ Mathematical Programming (Optimization Problem) Linear Programming(线性规划) Nonlinear Programming(非线性规划) Unconstrained Problems(无约束最优化方法) Constrained Problems(约束最优化方法) ☆ Variational Inequality (VI)(变分不等式) Monotone VI Non-monotone VI * Optimization Problem (Mathematical) Optimization Problem optimal solution(最优解) x* has smallest value(最小值) of f among all vectors(矢量) that satisfy the constraints(满足约束条件). (优化变量) (目标函数) (约束函数) * Optimization Problem: Examples Portfolio optimization(投资组合优化) variables: amounts invested in different assets constraints: budget, max./min. investment per asset, min. return o