## 数学代写|凸优化作业代写Convex Optimization代考|MATH3204

2022年9月27日

couryes-lab™ 为您的留学生涯保驾护航 在代写凸优化Convex Optimization方面已经树立了自己的口碑, 保证靠谱, 高质且原创的统计Statistics代写服务。我们的专家在代写凸优化Convex Optimization代写方面经验极为丰富，各种代写凸优化Convex Optimization相关的作业也就用不着说。

• Statistical Inference 统计推断
• Statistical Computing 统计计算
• (Generalized) Linear Models 广义线性模型
• Statistical Machine Learning 统计机器学习
• Longitudinal Data Analysis 纵向数据分析
• Foundations of Data Science 数据科学基础
couryes™为您提供可以保分的包课服务

## 数学代写|凸优化作业代写Convex Optimization代考|The Generalization to the Multi-Objective Case

Recently several papers have been published which propose multi-objective optimization algorithms that generalize single-objective optimization algorithms based on statistical models of objective functions [53, 101, 105, 106, 142, 224, 252]. The numerical results included there show the relevance of the proposed algorithms to the problems of multi-objective optimization with black-box expensive objectives. We present here a new idea for constructing relevant algorithms.

A multi-objective minimization problem can be stated almost identically to the single-objective problem considered in the previous subsection:
$$\min _{\mathbf{x} \subset \mathbf{A}} \mathbf{f}(\mathbf{x}), \mathbf{f}(\mathbf{x})=\left(f_1(\mathbf{x}), f_2(\mathbf{x}), \ldots, f_m(\mathbf{x})\right)^T, \mathbf{A} \subset \mathbb{R}^d$$ however, the concept of solution in this case is more complicated. For the definitions of the solution to a multi-objective optimization problem with nonlinear objectives, we refer to Chapter 1 .

In the case of multi-objective optimization, a vector objective function $\mathbf{f}(\mathbf{x})=$ $\left(f_1(\mathbf{x}) \cdot f_2(\mathbf{x}) \ldots \ldots f_m(\mathbf{x})\right)^T$ is considered. The same arguments, as in the case of single-objective optimization, corroborate the applicability of statistical models. The assumptions on black-box information and expense of the objective functions together with the standard assumptions of rational decision making imply the acceptability of a family of random vectors $\Xi(\mathbf{x})=\left(\xi_1(\mathbf{x}), \ldots, \xi_m(\mathbf{x})\right)^T, x \in \mathbf{A}$, as a statistical model of $\mathbf{f}(\mathbf{x})$. Similarly, the location and spread parameters of $\xi_i(\mathbf{x})$, denoted by $m_i(\mathbf{x}), s_i(\mathbf{x}), i=1, \ldots, r$, are essential in the characterization of $\xi_i(\mathbf{x})$. For a more specific characterization of $\Xi(\mathbf{x})$, e.g., by a multidimensional distribution of $\Xi(\mathbf{x})$, the available information usually is insufficient. If the information on, e.g., the correlation between $\xi_i(\mathbf{x})$ and $\xi_j(\mathbf{x})$ were available, the covariance matrix could be included into the statistical model. However, here we assume that the objectives are independent, and the spread parameters are represented by a diagonal matrix $\Sigma(\mathbf{x})$ whose diagonal elements are equal to $s_1, \ldots, s_m$. Similarly to the case of single-objective optimization, we assume that the utility of choice of the point for the current computation of the vector value $\mathbf{f}(\mathbf{x})$ has the following structure
$$v_{n+1}(\mathbf{x})=V_{n+1}\left(\mathbf{m}(\mathbf{x}), \Sigma(\mathbf{x}), \mathbf{y}^n\right),$$
where $\mathbf{m}(\mathbf{x})=\left(m_1(\mathbf{x}), \ldots, m_m(\mathbf{x})\right)^T$, and $\mathbf{y}^n$ denotes a vector desired to improve.

## 数学代写|凸优化作业代写Convex Optimization代考|Methodological Problems

Theoretical and algorithmic achievements of multi-objective optimization have implied also the expansion of respective applications. Among applied multiobjective optimization problems, expensive multimodal black-box problems are rather frequent. However, they constitute still a relatively little researched subfield of multi-objective optimization and deserve more attention from researchers. Since the statistical models based single-objective optimization algorithms well correspond to the challenges of single-objective global optimization of expensive black-box functions, they were generalized to the multi-objective case. As shown in the previous sections, the theoretical generalization is rather straightforward. Some experimental investigation was performed to find out how much the generalization corresponds to the expectations of their suitability for multi-objective optimization.
General methodological concepts of testing and comparison of mathematical programming algorithms and software are well developed; see [131]. The methodology, called Competitive Testing in [82], should normally be applied for the comparison of the well-established algorithms. This methodology is also extended for testing and comparison of multi-objective optimization algorithms; see, e.g., [42, $60,135,266,267]$. In the case of the well-researched classes of problems (e.g., convex multi-objective optimization), this methodology is universally applicable, only the selection of test functions should be specially selected taking into account the properties of the considered sub-class of problems, e.g., considered in $[66,129]$. The tests, based on special cases of real world applied problems, can be very useful for evaluating the efficiency of the respective algorithms; see, e.g., [154] where multi-objective portfolio problems are used for testing the algorithms aimed to distribute solutions uniformly in the Pareto optimal set.

However, the standard testing methodology is not well suitable for the algorithms considered in this chapter. The first difficulty is caused by the main feature of the targeted problems: they are supposed to be expensive. Therefore, a solution, found by an optimization algorithm applied, normally is rather rough. An optimization algorithm is as much useful as much its application aids a decision maker in making a final decision in the conditions of uncertainty reduced because of the application of the algorithm. The quantitative assessment of such a criterion of an algorithm is difficult.

## 有限元方法代写

tatistics-lab作为专业的留学生服务机构，多年来已为美国、英国、加拿大、澳洲等留学热门地的学生提供专业的学术服务，包括但不限于Essay代写，Assignment代写，Dissertation代写，Report代写，小组作业代写，Proposal代写，Paper代写，Presentation代写，计算机作业代写，论文修改和润色，网课代做，exam代考等等。写作范围涵盖高中，本科，研究生等海外留学全阶段，辐射金融，经济学，会计学，审计学，管理学等全球99%专业科目。写作团队既有专业英语母语作者，也有海外名校硕博留学生，每位写作老师都拥有过硬的语言能力，专业的学科背景和学术写作经验。我们承诺100%原创，100%专业，100%准时，100%满意。

## MATLAB代写

MATLAB 是一种用于技术计算的高性能语言。它将计算、可视化和编程集成在一个易于使用的环境中，其中问题和解决方案以熟悉的数学符号表示。典型用途包括：数学和计算算法开发建模、仿真和原型制作数据分析、探索和可视化科学和工程图形应用程序开发，包括图形用户界面构建MATLAB 是一个交互式系统，其基本数据元素是一个不需要维度的数组。这使您可以解决许多技术计算问题，尤其是那些具有矩阵和向量公式的问题，而只需用 C 或 Fortran 等标量非交互式语言编写程序所需的时间的一小部分。MATLAB 名称代表矩阵实验室。MATLAB 最初的编写目的是提供对由 LINPACK 和 EISPACK 项目开发的矩阵软件的轻松访问，这两个项目共同代表了矩阵计算软件的最新技术。MATLAB 经过多年的发展，得到了许多用户的投入。在大学环境中，它是数学、工程和科学入门和高级课程的标准教学工具。在工业领域，MATLAB 是高效研究、开发和分析的首选工具。MATLAB 具有一系列称为工具箱的特定于应用程序的解决方案。对于大多数 MATLAB 用户来说非常重要，工具箱允许您学习应用专业技术。工具箱是 MATLAB 函数（M 文件）的综合集合，可扩展 MATLAB 环境以解决特定类别的问题。可用工具箱的领域包括信号处理、控制系统、神经网络、模糊逻辑、小波、仿真等。