Systematic strategy for constructing a feature space if features are are naturally numerical
General idea
If you are using th-order basis (where is a positive integer), you include a feature for every possible product of different dimensions in your original input.
Example: XOR problem using a polynomial basis as a feature transformation
XOR with feature transformation
Let’s say for the XOR problem in 2 dimensions.
The feature transform is:
After training a perceptron for 4 iterations, we have: which corresponds to
Plotting this, we can see the gray shaded region classified as negative and the white region
For more complex dataset, the perceptron can still find appropriate classifiers using this trick: