Skip to main content

Table 4 Test results of ten neural networks, trained on 146 genes to diagnose 96 donors for DLBCL lymphoma

From: Neural network analysis of lymphoma microarray data: prognosis and diagnosis near-perfect

DC

NC

DC

NC

DC

NC

1.0000

0.9714

1.0000

0.9877

1.0000

0.9881

1.0000

0.9744

1.0000

0.9889

1.0000

0.9877

1.0000

0.9645

1.0000

0.9754

1.0000

0.9877

1.0000

0.9778

1.0000

0.9847

1.0000

0.9876

1.0000

0.9784

1.0000

0.9715

1.0000

0.9862

0.0000

0.0202

0.0000

0.0183

0.0000

0.0138

0.0000

0.0222

0.0000

0.0186

0.0000

0.0145

0.0000

0.9698*

0.0000

0.0176

0.0000

0.0153

0.0000

0.3162

0.0000

0.0182

0.0000

0.0140

0.0000

0.0170

0.0000

0.0205

0.0000

0.0155

1.0000

0.9820

1.0000

0.9870

1.0000

0.9849

1.0000

0.9817

1.0000

0.9849

1.0000

0.9839

1.0000

0.9821

1.0000

0.9790

1.0000

0.9858

1.0000

0.9803

1.0000

0.9859

1.0000

0.9849

1.0000

0.9611

1.0000

0.9811

1.0000

0.9841

0.0000

0.0127

0.0000

0.0276

0.0000

0.0217

0.0000

0.0415

0.0000

0.0435

0.0000

0.0221

0.0000

0.0109

0.0000

0.2255

0.0000

0.0213

0.0000

0.0115

0.0000

0.1417

0.0000

0.0216

0.0000

0.0115

0.0000

0.1463

0.0000

0.0218

1.0000

0.9857

1.0000

0.9703

1.0000

0.9953

1.0000

0.9850

1.0000

0.9765

1.0000

0.9953

1.0000

0.9877

1.0000

0.9752

1.0000

0.9953

1.0000

0.9436

1.0000

0.9742

1.0000

0.9644

1.0000

0.9861

1.0000

0.9760

1.0000

0.9950

0.0000

0.0149

0.0000

0.2234

1.0000

0.9817

0.0000

0.0169

0.0000

0.1811

0.0000

0.0005

0.0000

0.0166

0.0000

0.0336

0.0000

0.0005

0.0000

0.0153

0.0000

0.0177

0.0000

0.0006

0.0000

0.0149

0.0000

0.0152

0.0000

0.0005

    

0.0000

0.0005

    

0.0000

0.0177

    

0.0000

0.0180

    

0.0000

0.0177

    

0.0000

0.0184

    

0.0000

0.0218

  1. 1.0 indicates positive for DLBCL. DC indicates the actual donor class, with 0.0 being negative. NC gives the network evaluation. The NC cutoff throughout this work is: ≥ 0.50 is taken as 1 and <0.50 is taken as 0. * marks the errors