Data—Not Neatly Separable
Discover how to approach non-linearly separable data sets that challenge simple models like the perceptron. Learn why a single linear boundary often fails and explore the foundations for building multilayer neural networks to classify complex data more accurately using Python and scikit-learn.
We'll cover the following...
We'll cover the following...
More realistic movie dataset
Till now, we have used a simpler dataset. Simpler in one sense. The two classes of movies were actually linearly separable, i.e., you could draw a line in the space between the two classes and neatly separate the input space into two distinct regions. No data point or instance from the entire dataset violated this linear boundary. This worked really well for our first machine learning model—the perceptron, or a single neuron.
Linearly separable data
...