40 Algorithms Every Programmer Should Know

Voice categorization

Unsupervised learning can be used to classify individual voices in a voice file. It uses the fact that each individual's voice has distinct characteristics, creating potentially separable audio patterns. These patterns can then be used for voice recognition—for example, Google uses this technique in their Google Home devices to train them to differentiate between different people's voices. Once trained, Google Home can personalize the response for each user individually.

For example, let's assume that we have a recorded conversation of three people talking to each other for half an hour. Using unsupervised learning algorithms, we can identify the voices of distinct people in this dataset. Note that through unsupervised ...