Artificial Neural Network to Predict Varicocele Impact on Male Fertility through Testicular Endocannabinoid Gene Expression Profiles

Figure 7

ANN learning process. The picture summarizes the ANN learning process automatically performed after the introduction of ECS gene expression levels (inputs) and mean reproductive outcomes (outputs). As shown in the first column of keys, this analysis was carried out by taking into account the datasets (paired inputs/outputs) obtained from all the experimental animals (n = 35) as training examples. During 793 learning cycles the ANN displayed a progressive decrease of the training errors that reached, at the end of the learning step, a minimum, maximum, and average prediction error of 0.000000, 0.215529, and 0.009992, respectively. Red line: maximum error. Green line: average error. Blue line: minimum error.