Advances in Artificial Neural Systems / 2010 / Article / Tab 4 / Research Article
Comparison of Artificial Neural Network with Logistic Regression as Classification Models for Variable Selection for Prediction of Breast Cancer Patient Outcomes Table 4 Variables selections for Logistic Regression and Neural Network approaches.
(a) Disease-free survival. Variables NN LR Invaded nodes 100% X Clinical size stage 100% X Number nodes invaded 100% X SBR grade 98% X Histology 98% Necrosis 98% Oestrogen receptors 74% X Skin infiltrating tumour 21% X
(b) Mortality from cancer causes. Variables NN LR Clinical number nodules 100% Progesterone receptor 100% X Number nodes invaded 100% X Clinical size stage 98% X SBR grade 98% X Histology 98% Invaded nodes 98% Skin embolus 8% X
(c) Local recurrence. Variables NN LR Number nodes invaded 98% X Lymphatic embolus 95% X Ganglion invaded 95% Necrosis 95% X Oestrogen receptors 95% X Histology 48% X Number tumours 15% X Skin invasion 13% X
(d) Metastatic recurrence. Variables NN LR Clinical Size stage 100% X Invaded nodes 100% Progesterone receptors 100% X Number of nodes invaded 100% X SBR grade 98% X Histology 98% Oestrogen receptors 85% X