Unit 8: Data Mining II – Clustering Techniques
This unit extends the material in the previous unit to clustering techniques, which are useful for creating pattern classification models where the input classes are unknown (which we call unsupervised learning). When you finish this unit, you will be able to create programs capable of training and testing unsupervised learning models. As in the previous unit, we will implement these techniques using the scikit-learn module.
The clustering of input feature vectors can be accomplished in several different ways. This unit focuses on two techniques: K-means, which requires some knowledge of the number of classes, and hierarchical clustering, which allows the input data to gradually define the number of classes. Both methodologies have their place within the field of data science.
Completing this unit should take you approximately 5 hours.
8.1: Unsupervised Learning
8.2: K-means Clustering
8.3: Hierarchical Clustering
8.4: Training and Testing
Unit 8 Assessment
- Receive a grade