Clustering

Categorization using human-understandable names is difficult and not efficient. For an overview of common categorization in social sciences, see Social Categorization, Social group.

Now we can use learning algorithms to automatically group statistically distinguishable behavior clusters, the so-called behaviotypes. Many areas and people have been using this new technique, such as Prof. Ram Rajagopal, and neural-behavioral maps of larvae.

Cluster analysis is used in many disciplines to group objects according to a defined measure of distance.

• k-means, streaming k-means
• Gaussian mixture
• Power iteration clustering (PIC)
• Latent Dirichlet allocation (LDA)

Multi-scale unsupervised structure learning

Iterative Denoising Trees (IDT) algorithm:

1. Interpoint dissimilarity matrix construction: dissimilarity between a pair of multivariate time series
2. Embedding: embed comparison matrix into Euclidean space using multidimensional scaling (MDS)
3. Iteration; Convergence Checking
   1. Dimension Selection
   2. Clustering: hierarchical clustering, Bayesian Information Criterion

Gaussian mixture model-based clustering

mclust - R package for model-based clustering

---

🏷 Category=Computation Category=Machine Learning