Unsupervised learning is a set of techniques that helps discover patterns in unlabeled data. For example, suppose you're looking at dogs - some dogs are tall, others may have long fur. Although you're not exactly sure what they're called, you may notice some repeating patterns and similarities. Without knowing what is a "Greyhound", or "German Shepherd" you've already learned their important attributes! This is one type of unsupervised learning, called clustering. Unsupervised learning is powerful tool for detecting anomolies, finding useful features, and visualizing the structure of unlabeled data. In this workshop we will give you an introduction to unsupervised learning techniques in machine learning. After a crash course on unsupervised learning, we will cover a few strategies used by popular algorithms, including K-means and PCA. During the workshop, we will be using a popular Python machine learning library, scikit-learn.

Topics:

-Supervised vs. Unsupervised learning

-Implementation of unsupervised learning algorithms (PCA, k-means, and others)

-Applications of unsupervised learning (Anomaly detection, compression, visualization)

-Examples with scikit-learn

Duration: 2 hours.