在开放源码数据挖掘软件Weka(在爪哇写作)中,当我运行一些数据挖掘算法,如Linear回归Weka返回模型和测试数据的一些模型评价指标时。
看起来是这样的:
Correlation coefficient 0.2978
Mean absolute error 15.5995
Root mean squared error 29.9002
Relative absolute error 47.7508 %
Root relative squared error 72.2651 %
“相对绝对误差”和“Rooot相对正方形误差”的公式是什么?我无法理解这一点。我想用这些衡量尺度来评估我自己在Matlab的算法。
从"http://www.liacs.nl/home/joost/DM/mod_11_eval_lift_cost.ppt" rel=“norefererr”>这个 演示文稿,在幻灯片22中,并引用Wighten,这里是公式:
Relative absolute error
Root relative squared error
与
根相对平方错误的公式实际上是相对平方错误的公式。您需要选择此公式的平方根才能获得 Weka 输出的内容 。
Witten Weka书第177页,Ian H.,Eibe Frank和Mark A. Hall,“实用机器学习工具和技术”。 Morgan Kaufmann(2005年):578,相对方形错误的定义如下:
“错误是相对于使用简单的预测器时的错误。 有关的简单预测器仅仅是从 < 坚固 > 训练 < /坚固 > 数据中得出的实际值的平均值。 因此,相对的平方错误取总平方错误,并通过除以默认预测器的总平方错误来使其正常化。”
这与Weka执行一致,因此,人们需要计算所有相对错误的 " 强力 " 列车/强力 " 设定目标的平均值。
Which forums you are using for data mining questions? SO is mainly intended for programming, not for DM questions.
I m working on a project at the moment where I need to pick out the most common phrases in a huge body of text. For example say we have three sentences like the following: The dog jumped over the ...
I d like to find patterns and sort them by number of occurrences on an HEX file I have. I am not looking for some specific pattern, just to make some statistics of the occurrences happening there and ...
Does anybody know about any dataming libraries for .net?
I m planning to develop program in Java which will provide diagnosis. The data set is divided into two parts one for training and the other for testing. My program should learn to classify from the ...
I m using the explorer feature of Weka for classification. So I have my .arff file, with 2 features of NUMERIC value, and my class is a binary 0 or 1 (eg {0,1}). Sample: @RELATION summary @...
I ve got a somewhat ugly field in a database which holds the names of locations. For instance, Madison Square Gardens which has also been entered as "The Madison Square Gardens", etc. etc. I m ...
This is intended as a question quite open to any suggestions, hints or pointers. I wish to start playing around with home brewed automated investment models, the beginnings of which I have concepts ...