Table 2 Parameters of the simulated populations according to each tested scenario.
From: Head-to-head comparison of clustering methods for heterogeneous data: a simulation-driven benchmark
1 | 2 | 3 | 4 | 5 | 6 | 7 | |
|---|---|---|---|---|---|---|---|
General parameters | |||||||
Population size | 300/600/1200 | 300 | 300 | 300 | 300 | 300 | 300 |
Number of clusters | 6 | 2/6/10 | 6 | 6 | 6 | 6 | 6 |
Continuous variables | |||||||
Total number | 4 | 4 | 2/4/8 | 4 | 4 | 4 | 10 |
Proportion of relevant variables | 100% | 100% | 100% | 100% | 100% | 100% | 20%/50%/90% |
Degree of relevance | Mild | Mild | Mild | Low/mild/high | Mild | Mild | Mild |
Categorical variables | |||||||
Total number | 4 | 4 | 2 / 4 / 8 | 4 | 4 | 10 | 4 |
Proportion of relevant variables | 100% | 100% | 100% | 100% | 100% | 20%/50%/90% | 100% |
Degree of relevance | Mild | Mild | Mild | Mild | Low/mild/high | mild | Mild |
