Unsupervised learning relies more on artificial intelligence and less on human intelligence. In this video, learn how to determine when unsupervised learning is most useful.
- [Instructor] In machine learning, you aren't always tryna predict the value. Sometimes your goal is to find some structure in your dataset. Unsupervised learning is when you train an algorithm without giving it the answers for examples in your dataset. In the context of psychic learn, this means that you only provide a features matrix when you fit your algorithm. A features matrix is a two-dimensional grid of data where rows represent samples and columns represent features. Unlike supervised learning, there's no target factor. It's important to emphasize that unsupervised algorithms don't make predictions from the data. There are two common types of unsupervised learning algorithms. The first is clustering. Clustering is often used to discover natural groupings in a dataset, when common use is for market segmentation. Companies often have large amounts of customer information. By clustering customers into different segments, they can more efficiently sell or market to their customers. Another common type of unsupervised learning is dimensionality reduction. You can think of dimensionality reduction techniques as data compression algorithms. They can make your data take up less space on your computer. Having less features in your data can make visualizing your data easier as well speed up the fitting of your machine learning algorithms. So that's it. Unsupervised learning helps you discover structure in your data.
This course was created by Madecraft. We are pleased to host this content in our library.
- Why use scikit-learn?
- Supervised vs. unsupervised learning
- Linear and logistic regression
- Decision trees and random forests
- K-means clustering
- Principal component analysis (PCA)