Original Paper: Unsupervised Machine Learning for Clustering in Political and Social Research
Header Photo: From Unsplash
Setting the Stage for Clustering
Clustering can be thought of in two main ways: subdivide the data or not.
Subdividing the data means to partition the data into smaller clusters.
Opposite of partitioning or subdividing the data is to cluster data in a pairwise fashion: recursively from the “top-down” or from the “bottom-up”.
Prior to fitting clustering algorithms, we must first set the stage. This typically consists of assessing clusterability and then defining distance (similarity and dissimilarity between observations).
Assessing Clusterability
FEATURED TAGS
Computer Vision
Colorization
Deep Learning
Papers
Technical Writing
Computer Graphics
Tutorials
Mathematics
Graph Neural Network
Biomedical
Natural Language Processing
Graph Theory
Neural Network
Reinforcement Learning
Markov Decision Process
Paper Reading
Games
GVGAI
VGDL
MCTS
Clustering
Unsupervised Machine Learning
Social Science
