see for more details, e.g., [5] and [6]), and animated sensitivity analysis that provides the insight into the behavior of the solution produced by an algorithm (for more details, see, e.g., [7] and [8]). The review covers special algorithms and tools. However, other visualization approaches (e.g., projection approaches [9], worlds within worlds [10], parallel coordinates [11]) may also be adapted to animated usage seeking the goals of the abovementioned ways. In this paper, we present a new way of visual and animated analysis of extremal problems. It allows us to discover new properties of the objective function and to search for a new coordinate system (i.e. for a linear transformation of the system of variables). In fact, a new phenomenon that characterizes the extremal problem has been discovered. Its analysis, presented in the paper, raises more items for further discussions and research rather than gives answers. The paper tries to reveal fields of application of this phenomenon. However, a slight relationship with practical applications has been discovered. 2. The Aim of Analysis 3. Information Structures ......
摘要:本文的结果在于建立了一种分析极值问题新方法的基础。发现了一种刻画极值问题的新现象。基于问题中目标函数观测集的交互性的知识发现，构造了动画视觉分析的方法。该分析的目的是在定义域中找到一个方向，从而最大化在该方向随机选择的两个点上的目标函数值的平均绝对差，或者（并且）最大化每个距离单元上该方向随机选择的两个点上的目标函数值的平均绝对差。提出的方法需要生成许多数据集合。有时这种生成是计算昂贵的。因此，这篇文章讨论的的思想可以应用于侦查员不但想解决极值问题而且想得到更多信息的的情形。
1. 引言
2.分析的目的 ......
