宣告： 只作為自己閱讀論文的相關筆記記錄，理解有誤的地方還望指正 

論文下載連結：

概念：資料驅動？

Solving evolutionary optimization problems driven by data collected in simulations, physical experiments, production processes, or daily life are termed data-driven evolutionary optimization.

現象：

Most existing research on EAs is based on an implicit assumption that evaluating the objectives and constraints of candidate solutions is easy and cheap. However, such cheap functions do not exist for many real-world optimization problems.

EA盛行基於假設評估目標以及候選解約束條件簡單而且評估的代價低廉，問題就是它只是現實的一種近似逼近，以及應用起來評估的代價還是挺高的。

 資料驅動中資料難點

文章的目的：

This paper aims to provide an overview of recent advances in the emerging research area of data-driven evolutionary optimization 

  文章的整體 框架：

資料驅動進化演算法的主要的組成

資料的分類：

Off-line 的存在的問題以及相應的處理方法

而對於On-line Data-driven Optimization Methodologies的關鍵點以及難點就是：

怎麼樣取樣？ 針對新的資料進行取樣？新的資料主要的目的是用來進行一步步的修訂模型的，若是取樣不當很可能使得構建的模型在結構上發生變化  

ON-line Data-drive methodologes sample

Promising samples are located around the optimum of the surrogate model, and the accuracy of the surrogate model in the promising area is enhanced once the promising solutions are sampled [8], [14].

• Uncertain samples are located in the search space where the surrogate model is likely to have a large approximation error and has not been fully explored by the EA 

[8] Y. Jin, M. Olhofer, and B. Sendhoff, “On evolutionary optimization with approximate fitness functions,” in Proceedings of the Genetic and Evolutionary Computation Conference. Morgan Kaufmann Publishers Inc., 2000, pp. 786–793.

[14] Y. Jin, M. Olhofer, and B. Sendhoff, “A framework for evolutionary optimization with approximate fitness functions,” IEEE Transactions on Evolutionary Computation, vol. 6, no. 5, pp. 481–494, 2002.