A Receiver Operating Characteristic (ROC) curve is a graphical visual that illustrates the predictive ability of a binary classifier system. The ROC curve is created by plotting a graph of the TPR against the False Positive Rate (FPR) at various threshold settings. This gives us Sensitivity versus (1 - Specificity). A ROC curve typically looks like this: 

After acquiring the necessary skills, we are ready to analyze in detail the algorithms used for building the neural networks.