3 basic types of machine learning algorithms:
supervised, unsupervised and reinforcement learning.
Example: Predicting the price of a house.
Supervised learning works with a dataset made of features and labels. If we take a dataset of different houses which features are:
- Distance from the city
- Type of house
- Number of rooms
And labels are the price of each house.
We start by fitting the classifier with these features and labels to create a predictive model. Once this model is trained, we can use a new input (new house), run the predictive model with it and get a prediction of a price for that new house.
Unsupervised learning works a little differently as instead of predicting a particular price, we are trying to identify patterns.
If we take the same dataset as before, the only difference is that we do not use labels to train the predictive model. As no label is provided, the output will be clusters of house that have similar features so we can get an indication of how much a new house is going to be based on similar house.