Overview
Jay Magidson is the founder of Statistical Innovations Inc. and a developer of applied latent class and finite-mixture methodology, particularly for marketing and market segmentation. With Jeroen K. Vermunt he developed the Latent GOLD software and popularized the LC factor model and the use of LC clustering as a probabilistic alternative to K-means.
Key Contributions / Features
- Co-author of Magidson-Vermunt (2004) Latent Class Models, the unified cluster/factor/regression treatment of LC modeling.
- Co-developer of Latent GOLD (with Jeroen K. Vermunt).
- Advocate of LC cluster analysis as a model-based improvement over K-means and of LC factor models with biplot displays for marketing segmentation.
Related