Skip to main content

Table 5 Maize datasets

From: Genome-enabled prediction using probabilistic neural network classifiers

Upper class
  MLP15% PNN15% MLP30% PNN30%
GY-HI 0.235 (0.126) 0.306 (0.118) 0.429 (0.108) 0.509 (0.102)
GY-LO 0.168 (0.065) 0.188 (0.076) 0.358 (0.107) 0.408 (0.107)
GY-SS 0.199 (0.093) 0.204 (0.110) 0.363 (0.111) 0.453 (0.119)
GY-WW 0.239 (0.131) 0.382 (0.175) 0.410 (0.117) 0.477 (0.111)
Middle class
  MLP40% PNN40% MLP70% PNN70%
ASI-SS 0.465 (0.096) 0.495 (0.092) 0.724 (0.076) 0.746 (0.074)
ASI-WW 0.436 (0.091) 0.481 (0.088) 0.706 (0.072) 0.722 (0.084)
Lower class
  MLP15% PNN15% MLP30% PNN30%
FFL-SS 0.185 (0.087) 0.288 (0.137) 0.383 (0.106) 0.465 (0.096)
MFL-SS 0.205 (0.101) 0.343 (0.149) 0.421 (0.119) 0.499 (0.112)
FFL-WW 0.197 (0.102) 0.298 (0.161) 0.413 (0.120) 0.506 (0.133)
MFL-WW 0.199 (0.094) 0.288 (0.155) 0.437 (0.133) 0.516 (0.139)
GLS-1 0.269 (0.096) 0.338 (0.135) 0.476 (0.096) 0.526 (0.092)
GLS-2 0.320 (0.140) 0.447 (0.157) 0.524 (0.101) 0.642 (0.093)
GLS-3 0.372 (0.138) 0.456 (0.149) 0.496 (0.128) 0.589 (0.116)
GLS-4 0.350 (0.135) 0.487 (0.147) 0.439 (0.110) 0.512 (0.111)
GLS-5 0.161 (0.072) 0.208 (0.107) 0.429 (0.098) 0.538 (0.118)
GLS-6 0.320 (0.091) 0.400 (0.109) 0.431 (0.094) 0.491 (0.098)
  1. Mean values of the area under the precision-recall curve AUCpr AUCpr (standard deviation in parentheses) of 50 random partitions for 15 and 30 % upper classes for grain yield (GY) in four environments (HI, LO, SS, and WW), for 40 and 70 % middle class for anthesis-silking interval (ASI) in two environments (SS and WW), and for 15 and 30 % lower classes for four traits, female flowering (FFL) and male flowering (MFL) in two environments (SS and WW); for gray leaf spot resistance (GLS) in six environments (1–6) and for classifiers MLP and PNN. Numbers in bold are the highest AUCpr values between MLP and PNN for 15 and 30 %