Jeroen Janssens • 11/14/2013

Stochastic Outlier Selection

This technical blog article by Jeroen Janssens explains Stochastic Outlier Selection (SOS), a machine learning algorithm for outlier detection. It details how SOS works by transforming data into a dissimilarity matrix, computing affinities and binding probabilities, and outputting an outlier probability for each point. The article discusses its relation to t-SNE, its perplexity parameter, and includes a link to a Python implementation.

0 comments

#Python #Machine Learning #Outlier Detection