决策连续形式背景的可视化数据离散化方法

连续形式背景离散化是形式概念分析领域重要的基础问题之一。本文针对形式背景离散化的特殊要求，提出了一种可视化的数据离散化方法。该方法借助可视化方法对数据类别分布进行表示，将连续数据分布转化为图形分布，进一步利用视觉模糊性对图形空间进行处理，进而将决策连续背景离散化。通过UCI数据集上的实验表明，与传统离散化方法相比，采用本文方法进行数据离散化后的二值形式背景具有结构简单且不失准确性的优点。

Visual discretization for decision continuous formal context

Discretization for Decision Continuous Formal Context is one of basic issues in the field of formal concept analysis. According to the special requirements of formal concept analysis field for data discretization, this paper presents a method which is in the form of visualization. This method present the data distribution by visualization, convert the data distribution to figure distribution, and fuzzy analysis the figure to discretize the decision continuous formal context. The whole progress considers not only the category distribution, but also the fuzzy of the data. The experiments show that compared to the traditional discretization methods, the binary formal context after using this method to discretize data has the advantages of simple structure and without loss of accuracy.