Table 1.

Results of simulation scenarios. Proportions of 500 simulation runs in which |$X_1$| is correctly selected at the first split in scenarios 1–3 and 5–8 and |$X_1$| then |$X_2$| and |$X_3$| are correctly selected as the first three splits in scenarios 4 and 9. |$n$| is the total sample size, and |$p$| denotes the total number of candidate split variables. In scenarios 6–9, the sample size is 300.

  |$n=250$||$n=500$|
ScenarioMethod|$p=20$||$p=50$||$p=20$||$p=50$|
 Weighted0.0760.0400.1360.080
1. Non-tree, exponentialSimple0.4220.2620.5860.466
 DIPM0.4460.2900.6000.476
 Weighted0.1160.0660.1920.076
2. Non-tree, WeibullSimple0.5600.4320.7240.664
 DIPM0.5660.4440.7500.682
 Weighted0.7000.4980.9120.856
3. Tree of depth 2Simple0.7320.6500.9100.874
 DIPM0.7480.6800.9100.888
 Weighted0.7400.6380.9500.946
4. Tree of depth 3Simple0.7600.7580.9100.902
 DIPM0.7840.7840.9420.918
 Weighted0.0720.0360.0680.040
5. Non-tree, non-PHSimple0.1460.0620.2320.134
 DIPM0.1560.0660.2420.124
 No. of Z Vars.Weighted MethodSimple Cox splitsDIPM Method 
 10.5860.6440.694 
6. Non-tree, exponential100.0200.0980.070 
 1000.0000.0040.002 
 10.6420.7080.756 
7. Non-tree, Weibull100.0260.1300.110 
 1000.0000.0040.002 
 10.9300.9940.994 
8. Tree of depth 2100.6160.9700.962 
 1000.1260.8120.770 
 10.6340.6360.656 
9. Tree of depth 3100.0920.3020.270 
 1000.0020.0600.040 
  |$n=250$||$n=500$|
ScenarioMethod|$p=20$||$p=50$||$p=20$||$p=50$|
 Weighted0.0760.0400.1360.080
1. Non-tree, exponentialSimple0.4220.2620.5860.466
 DIPM0.4460.2900.6000.476
 Weighted0.1160.0660.1920.076
2. Non-tree, WeibullSimple0.5600.4320.7240.664
 DIPM0.5660.4440.7500.682
 Weighted0.7000.4980.9120.856
3. Tree of depth 2Simple0.7320.6500.9100.874
 DIPM0.7480.6800.9100.888
 Weighted0.7400.6380.9500.946
4. Tree of depth 3Simple0.7600.7580.9100.902
 DIPM0.7840.7840.9420.918
 Weighted0.0720.0360.0680.040
5. Non-tree, non-PHSimple0.1460.0620.2320.134
 DIPM0.1560.0660.2420.124
 No. of Z Vars.Weighted MethodSimple Cox splitsDIPM Method 
 10.5860.6440.694 
6. Non-tree, exponential100.0200.0980.070 
 1000.0000.0040.002 
 10.6420.7080.756 
7. Non-tree, Weibull100.0260.1300.110 
 1000.0000.0040.002 
 10.9300.9940.994 
8. Tree of depth 2100.6160.9700.962 
 1000.1260.8120.770 
 10.6340.6360.656 
9. Tree of depth 3100.0920.3020.270 
 1000.0020.0600.040 
Table 1.

Results of simulation scenarios. Proportions of 500 simulation runs in which |$X_1$| is correctly selected at the first split in scenarios 1–3 and 5–8 and |$X_1$| then |$X_2$| and |$X_3$| are correctly selected as the first three splits in scenarios 4 and 9. |$n$| is the total sample size, and |$p$| denotes the total number of candidate split variables. In scenarios 6–9, the sample size is 300.

  |$n=250$||$n=500$|
ScenarioMethod|$p=20$||$p=50$||$p=20$||$p=50$|
 Weighted0.0760.0400.1360.080
1. Non-tree, exponentialSimple0.4220.2620.5860.466
 DIPM0.4460.2900.6000.476
 Weighted0.1160.0660.1920.076
2. Non-tree, WeibullSimple0.5600.4320.7240.664
 DIPM0.5660.4440.7500.682
 Weighted0.7000.4980.9120.856
3. Tree of depth 2Simple0.7320.6500.9100.874
 DIPM0.7480.6800.9100.888
 Weighted0.7400.6380.9500.946
4. Tree of depth 3Simple0.7600.7580.9100.902
 DIPM0.7840.7840.9420.918
 Weighted0.0720.0360.0680.040
5. Non-tree, non-PHSimple0.1460.0620.2320.134
 DIPM0.1560.0660.2420.124
 No. of Z Vars.Weighted MethodSimple Cox splitsDIPM Method 
 10.5860.6440.694 
6. Non-tree, exponential100.0200.0980.070 
 1000.0000.0040.002 
 10.6420.7080.756 
7. Non-tree, Weibull100.0260.1300.110 
 1000.0000.0040.002 
 10.9300.9940.994 
8. Tree of depth 2100.6160.9700.962 
 1000.1260.8120.770 
 10.6340.6360.656 
9. Tree of depth 3100.0920.3020.270 
 1000.0020.0600.040 
  |$n=250$||$n=500$|
ScenarioMethod|$p=20$||$p=50$||$p=20$||$p=50$|
 Weighted0.0760.0400.1360.080
1. Non-tree, exponentialSimple0.4220.2620.5860.466
 DIPM0.4460.2900.6000.476
 Weighted0.1160.0660.1920.076
2. Non-tree, WeibullSimple0.5600.4320.7240.664
 DIPM0.5660.4440.7500.682
 Weighted0.7000.4980.9120.856
3. Tree of depth 2Simple0.7320.6500.9100.874
 DIPM0.7480.6800.9100.888
 Weighted0.7400.6380.9500.946
4. Tree of depth 3Simple0.7600.7580.9100.902
 DIPM0.7840.7840.9420.918
 Weighted0.0720.0360.0680.040
5. Non-tree, non-PHSimple0.1460.0620.2320.134
 DIPM0.1560.0660.2420.124
 No. of Z Vars.Weighted MethodSimple Cox splitsDIPM Method 
 10.5860.6440.694 
6. Non-tree, exponential100.0200.0980.070 
 1000.0000.0040.002 
 10.6420.7080.756 
7. Non-tree, Weibull100.0260.1300.110 
 1000.0000.0040.002 
 10.9300.9940.994 
8. Tree of depth 2100.6160.9700.962 
 1000.1260.8120.770 
 10.6340.6360.656 
9. Tree of depth 3100.0920.3020.270 
 1000.0020.0600.040 
Close
This Feature Is Available To Subscribers Only

Sign In or Create an Account

Close

This PDF is available to Subscribers Only

View Article Abstract & Purchase Options

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Close