weka--数据离散化

dengzong1707

于 2018-04-27 16:54:00 发布

阅读量2.1k

点赞数 1

文章标签： runtime

原文链接：http://www.cnblogs.com/zuimeiyujianni/p/8963335.html

版权

本文介绍了Weka库中的一个实例过滤器`weka.filters.unsupervised.attribute.Discretize`，该过滤器能够将数据集中的数值属性转换为名义属性。离散化过程可以通过等宽或等频方式进行。文章详细列出了该过滤器的各项参数选项，包括设置离散化的bin数量、是否生成二进制属性、控制实例权重和调整切点精度等。同时，还提到可以跳过分类属性并提供调试选项。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

NAME
weka.filters.unsupervised.attribute.Discretize

SYNOPSIS
An instance filter that discretizes a range of numeric attributes in the dataset into nominal attributes. Discretization is by simple binning. Skips the class attribute if set.

OPTIONS
spreadAttributeWeight -- When generating binary attributes, spread weight of old attribute across new attributes. Do not give each new attribute the old weight.

makeBinary -- Make resulting attributes binary.

debug -- If set to true, filter may output additional info to the console.

bins -- Number of bins.

doNotCheckCapabilities -- If set, the filter's capabilities are not checked before it is built. (Use with caution to reduce runtime.)

findNumBins -- Optimize number of equal-width bins using leave-one-out. Doesn't work for equal-frequency binning

attributeIndices -- Specify range of attributes