


Machine Learning models don't support such data natively.įortunately, with one-hot encoding, we can ensure that we can still use these features - simply by converting them into numeric vector format in a smart way. to illustrate the winners in some kind of competition. Neither is the case when your dataset contains an ordered list of names, e.g. For example, if your dataset contains categories, you have no numbers in your dataset. Now, when you look at this from a real-world point of view, you might get into a struggle soon when you look at datasets. It possibly makes the hype a little bit less fascinating, but it's the truth. That is, they are mathematical models which improve themselves by performing mathematical optimization. Machine Learning models work with numbers.
