crisp

Crisps the fuzzy and possibilistic membership degrees from the fuzzy or possibilistic clustering algorithms.

cluster

Partitioning clustering divides the objects in a data set into non-overlapping subsets or clusters by using the prototype-based probabilistic and possibilistic clustering algorithms. This package covers a set of the functions for Fuzzy C-Means (Bezdek, 1974) <doi:10.1080/01969727308546047>, Possibilistic C-Means (Krishnapuram & Keller, 1993) <doi:10.1109/91.227387>, Possibilistic Fuzzy C-Means (Pal et al, 2005) <doi:10.1109/TFUZZ.2004.840099>, Possibilistic Clustering Algorithm (Yang et al, 2006) <doi:10.1016/j.patcog.2005.07.005>, Possibilistic C-Means with Repulsion (Wachs et al, 2006) <doi:10.1007/3-540-31662-0_6> and the other variants of hard and soft clustering algorithms. The cluster prototypes and membership matrices required by these partitioning algorithms are initialized with different initialization techniques that are available in the package 'inaparc'. As the distance metrics, not only the Euclidean distance but also a set of the commonly used distance metrics are available to use with some of the algorithms in the package.

Zeynel Cebeci

ppclust

Probabilistic and Possibilistic Cluster Analysis

Figen Yildiz

Alper Tuna Kavlak

Cagatay Cebeci

Hasan Onder

crisp function

<dl> <dt>u</dt>
<dd>a numeric matrix containing the fuzzy or possibilistic membership degrees of the data objects.</dd> <dt>method</dt>
<dd>a string for selection of the crisping method. The default is max that assigns the data object to the cluster in which the object has maximum membership. The alternative is threshold that assigns the objects to a cluster if its maximum membership degree is greater than <code>tv</code>, a threshold value.</dd> <dt>tv</dt>
<dd>a number for the threshold membership degree. The default is 0.5 with the <code>method</code> is threshold if it is not speficied by the user.</dd></dl>

Arguments

Author

Crisp the fuzzy membership degrees — crisp

<dl>

 <dt>u</dt>
<dd>a numeric matrix containing the fuzzy or possibilistic membership degrees of the data objects.</dd>

 <dt>method</dt>
<dd>a string for selection of the crisping method. The default is max that assigns the data object to the cluster in which the object has maximum membership. The alternative is threshold that assigns the objects to a cluster if its maximum membership degree is greater than <code>tv</code>, a threshold value.</dd>

 <dt>tv</dt>
<dd>a number for the threshold membership degree. The default is 0.5 with the <code>method</code> is threshold if it is not speficied by the user.</dd>

</dl>

crisp: Crisp the fuzzy membership degrees

Description

Usage

Value

Arguments

Author

Details

Examples