machine learning - How to compute accuracy for cluster evaluation in Weka -

How do we calculate accuracy for groups using Weka?

I can use this formula:

  Accuracy (A) = (TP + TN) / total # samples  

But how do I get positive in positive, false positive, right negative and incorrect production in the Weka tool?

There are some different clustering modes in Weka:

Use (Default): After clustering, examples of training in the cluster developed by Waka are categorized and calculates the percentage of the examples coming in each cluster. For example, in Cluster 1 and X, X% in Cluster 1, etc.

Supplyed Test Set: It is possible that they evaluate clustering on different test data if the cluster representation is a probability like EM algorithms.

Clustering evaluation using classes: In this mode weka first ignores class attribute and generates clustering. During the trial, it provides class labels to groups based on the majority value of the class attribute within each cluster. Finally, this classification calculates the error and also shows related delusional matrix.