A Comparison Study on Rule Extraction from Neural Network Ensembles, Boosted Shallow Trees, and SVMs

Table 11

Comparison of rule extraction algorithms for the Breast Cancer classification problem, based on 10-fold cross-validation on single networks or ensembles of neural networks (last two rows).

Rule extraction technique	Rules Pred. Acc.	Avg. #rules	Avg. #Ant.
SSV (10 CV) [9]	96.3 (0.2)	3	–
FSM (10 CV) [9]	96.5	12	–
MINERVA (10 CV) [10]	94.5 (1.5)	4.2	3.3
NeuroLinear + GRG (10 CV) [11]	96.0	2	–
Re-RX + J48graft (10 × 10 CV) [12]	95.8 (1.6)	4.8	1.7
DIMLP-B (10 × 10 CV)	96.5 (0.3)	12.5	2.7