链接：
IEEE Xplore Search Results https://ieeexplore.ieee.org/search/searchresult.jsp?newsearch=true&queryText=Insights%20on%20Transfer%20Optimization:%20Because%20Experience%20is%20the%20Best%20Teacher
题目的信息：
（１）关于迁移优化的见解，内容应该涉及到对迁移优化的理解，以及提出自己的观点看法
（２）“Experience is the best Teacher”应该是一大堆的而案例说明先验知识的好处
我就是觉得看到论文说： 自己的见解，然后是强烈的阐述迁移优化为什么可行？然后是基于它的先验知识？？
希望自己觉都获得的：
（1） 主要是论文人家怎么写的？ 为什么一个“Insight”能够发表出来？？
（2） 迁移优化到底是什么？
（3） 怎么进一步的证明自己的论点“Experience is the best Teache” ？ 以及i怎么建模？ 怎样的数学表达式？？
摘要
Traditional : solvers tend to start the search from scratch by assuming zero prior knowledge about the task at hand
( 传统的倾向于： 从头开始入手，假设对自己需要求解的问题没有任何的先验知识。 就对应对该问题没有)没有任何的头绪In contrast： humans routinely make use of a pool of knowledge drawn from past experiences whenever faced with a new task. （显然作者是并不太同意这种： 实际中，我们总是会不自觉的应用自己的经验，当面对未知问题的时候）
This is often an effective approach in practice as real-world problems seldom exist in isolation. Similarly, practically useful artificial systems are expected to face a large number of problems in their lifetime, many of which will either be repetitive or share domain-specific similarities。
（1） 实际的问题之间是存在关联的，并不总是存在彼此无关的问题。
（2） 实际中有很多问题是具有相似性以及求解具有重复性。
所以
this paper sheds light on recent research advances in the field of global black-box optimization that champion the theme of automatic knowledge transfer across problems。
“theme ： automatic knwledge trasfer across problems----跨问题的自动知识迁移”
Introduction ：
立足的现象：
As a result, humans routinely resort to various information sources, including a pool of knowledge extracted from past
problem-solving experiences,when faced with a never before seen challenge or task.
句子：
（1） In many practical applications involving time sensitive actions and/or high cost of evaluations, ignoring the knowledge gained from previous optimization exercises can lead to deleterious computational overheads in the re-exploration of similar search spaces.
有害计算
（2） For the case of black-box optimization, where little problem-specific data is available beforehand, efforts in automatic knowledge transfer have been relatively rare; thereby establishing the need to devise new online approaches that can harness recurrent patterns between problem-solving exercises.
对于***，在**方面的研究还比较少，因此有建立***的必要
（3） manual intervention -----人工干预
（4）The practical motivation for incorporating such cognitive capabilities into optimization solvers is derived from the growing presence of modern technologies such as cloud computing and the Internet of Things (IoT), which enable large-scale storage and seamless information communication facilities.
将这种认知能力纳入优化求解者的实际动机来自于现代技术的日益增长。
（5）If we naively store the raw data in memory, not only is the extracted knowledge too coarse and possibly overfitting to the original problem
如果我们天真地**
（6）With the IoT giving rise to widespread inter-connectivity of physical devices and relatively easy access to diverse information streams, the present paper sheds light on the emerging scope of black-box optimization solvers to incorporate the general theme of transfer optimization.
文章的结构：
（7） On one hand, Bayesian optimization is extremely data efficient, but is exclusively a model-based
approach [19]. In contrast, nature-inspired techniques, albeit
less data efficient, provide considerable flexibility with the interplay of evolutionary mechanisms and model-based transfer[16], [24].
（8）With this in mind， the present paper attempts to provide a summary of the current state-of-the-art with a clear
exposition of the complementary nature of different research strands, so as to facilitate a unification of ideas leading to the
design of powerful transfer optimization engines in the future
考虑到这一点 互补性 以便得到统一的观点*
（9）a diverse array of noteworthy real-world examples are identified,covering topics such as machine learning,robotics, engineering design, etc
(10) Thereafter, we outline promising future directions that are expected to play a pivotal role in establishing automatic knowledge encoding and transfer mechanisms as intrinsic features of optimization search (Section VI). Finally, Section VII encapsulates the paper and presents concluding remarks
[16] L. Feng, Y. S. Ong, A. H. Tan, and I. W. Tsang, “Memes as building blocks: A case study on evolutionary optimization+ transfer learning for routing problems,” Memetic Comput., vol. 7, no. 3, pp. 159–180, 2015.
[24] A. Gupta, Y. S. Ong, and L. Feng, “Multifactorial evolution: Toward evolutionary multitasking,” IEEE Trans. Evol. Comput., vol. 20, no. 3, pp. 343–357, Jun. 2016
感想：
（1） 文章的逻辑思路
文章xia先给定义-----目前算法的贝叶斯以及自然启发式优化算法作为一个辅政------再说具体的实践中的例子----那些可以作为作者做简历理论的支撑---总结
总体的这个逻辑我是没有想到的，我停留在题目上的Insight 以为是大量的实验证明，通过实验忠明一种想法，仅仅的只想到实验，没有想到联系实际的问题以及实际的yi应用学科。
（2） 文章的句子
句子写的特别的好，有很多的句子的表达自己是没有见过的。整段的都想记下来
