d) The results of very high (100) and very low (0.0001) learning rates are poor. Both perceptrons where far from converging with errors greater of 0.9. On the other hand, middle ranged learning rates get a better result with precision of about \(10^{-2}\).