Machine Learning is a subfield of A.I. with three broad categories:
❇️ Supervised learning: A model is trained using example inputs for which the outcomes (labels) are known:
➡️ Predicting house prices based on environmental variables
➡️ Predicting which class an object in an image belongs to
➡️ Determining the sentiment of a piece of text through text classification
➡️ ...
❇️ Unsupervised learning: No labels are provided to the learning algorithm, so it has to look for structure in the input itself:
➡️ Customer segmentation
➡️ Automatic recommendations in webshops/Spotify/YouTube
➡️ Anomaly detection (fraud detection)
➡️ ...
❇️ Reinforcement learning: Training models to make a sequence of decisions based on feedback through "rewards" in an uncertain, potentially complex environment:
➡️ Optimising cooling in Google data centres
➡️ Self-driving cars
➡️ Automated buy/sell actions on the stock market
➡️ ...