Each one has some attributes and I want to classify them as good or bad customers. Since I don't have a training set (i.e. having for some IDs as their score of 0 or 1), how can I classify them into two groups? I understand that regression (logistic for example) cannot take place since I don't have a dependent variable. Could a solution be clustering for example and have only two clusters?
Yes, you can do clustering and can use the Mean Split Silhouette (MSS) as a measure of cluster heterogeneity and to estimate the number of significant clusters by choosing the number of clusters as the one that minimize the MSS and produces the most homogeneous clusters in the data.
Thanks majid and mahmoud , for ur concern i appreciate your views thanks
Yes, you can use the multivariate cluster data analysis with two means if you have more variables.
- Badges
- Science topic