Signing you in…

Categorical features: target encoding and pitfalls

Categorical features: how to explain "Moscow" to a neural network

Neural networks work with numbers. "Moscow", "Berlin", "Tokyo" are not numbers. The encoding task is to turn a category into a number without introducing a false order. If you tell the model Moscow=1, Berlin=2, Tokyo=3, it will infer that Tokyo is "three times larger" than Moscow. That is not true. Choosing the right encoding method has a serious impact on model quality.

Content is available with subscription.
Get full access to all courses on the platform for one year with a single payment.
Unlike other platforms that charge per course, here you get everything for one price, and after one year of use there will be no automatic charge for the following year.