= Generalized iterative scaling
{wiki=Generalized_iterative_scaling}
Generalized Iterative Scaling (GIS) is an algorithm used primarily in the context of statistical modeling and machine learning, particularly for optimizing the weights of a probabilistic model that adheres to a specified distribution. It is particularly useful for tasks involving maximum likelihood estimation (MLE) in exponential family distributions, which are common in various applications like natural language processing and classification tasks.
Back to article page