Clustering Data
summary
Covers several clustering algorithms and how to analyze their results: K-means and Mean Shift (centroid-based), DBSCAN and OPTICS (nearest-neighbor/density-based), and Spectral Clustering (embedding-based). Shows how to inspect clusterings using PCA and t-SNE embeddings and demonstrates these methods on multidimensional Telco Customer Churn and FIFA 2019 datasets.
AI-written summary · about
chapters (12)
- 0:00Data clustering intro
- 1:13What is clustering
- 3:43Cluster challenges
- 7:28Distances and affinities
- 13:55Clustering methods
- 17:35DBSCAN and k-means
- 25:43Evaluating partitions
- 34:23FIFA data prep
- 41:01Visualizing players
- 53:07FIFA clustering results
- 1:06:00Clustering vs classification
- 1:08:29Q&A
related talks
35mClustering: A Guide for the Perplexed
26mMathematical Approaches to Clustering
47mAssessing the quality of a clustering
34mDimensionally reducing data squeezing out the good stuff
25mThe Art of Clustering: The Good, The Bad and The Beautiful
34mHDBSCAN, Fast Density Based Clustering, the How and the Why
26mUnleash Big Data Clustering
51mTopological space creation and Clustering at BigData scale