Neural Networks with Python - [Part 4]

In this fourth tutorial on artificial neural networks with Python and neurolab, we are training the perceptron we've created in the previous tutorial.

In neurolab this is basically one line of code. We will have to call the 'train' method and pass it the following arguments:

• the data (features)
• the labels (target)
• the number of epochs to train it for - in this case 100
• 'show' which display the training status at a frequency of 'k' epochs - in this case 20
• and the learning rate - in this case 0.01

Then, we do a little bit of plotting to observe how the training goes and how the error is minimized.

One another thing we need to go into is to determine the performance of the perceptron on new data. For this we are using the 'sim' method in neurolab. After throwing 2 new (unseen) datapoints at the perceptron, we notice that it labeled them correctly.

Please watch the video below for the full walk-through and to possibly follow along with the code:

---

To stay in touch with me, follow @cristi

---

Cristi Vlad Self-Experimenter and Author