KI-Glossar
Das vollständige Wörterbuch der Künstlichen Intelligenz
Silhouette Coefficient
Clustering evaluation metric measuring intra-cluster cohesion relative to inter-cluster separation, ranging from -1 to 1.
DBSCAN Clustering
Density-based clustering algorithm that groups nearby points while marking points in low-density regions as outliers.
Similarity Matrix
Square structure containing similarity coefficients between all pairs of observations, essential for many clustering algorithms.
Divisive Clustering
Top-down approach to hierarchical clustering starting with all observations in a single cluster, recursively dividing it into sub-clusters.
K-modes Algorithm
Extension of K-means for categorical data using simple dissimilarity and mode as centrality measure instead of the mean.
Fuzzy Clustering
Method where each point can belong to multiple clusters with different membership degrees, unlike hard clustering where membership is binary.
OPTICS Algorithm
Extension of DBSCAN producing an ordering of points that reveals the density structure of data, allowing extraction of clusters at different densities.
Dunn Index
Clustering validation index measuring the ratio between the smallest inter-cluster distance and the largest intra-cluster diameter.
Clustering validation
Process of quantitative and qualitative evaluation of clustering results using internal, external or relative indices to measure quality.
Density-based clustering
Category of algorithms that identify clusters as dense regions separated by regions of low density in the data space.
Partitioning clustering
Family of algorithms directly dividing data into K non-hierarchical clusters, typically optimizing a distance or similarity criterion.