Fill in Missing Values in an Astronomy Dataset

SynthesizeMissingValues allows for missing data to be imputed based on an estimation of the underlying distribution of the dataset.

Obtain a dataset related to features of the moons of Jupiter that contains missing values.

SynthesizeMissingValues allows for missing values to be filled in by estimating a probability density function for the dataset using LearnDistribution. For this dataset, note that small-radius moons are imputed by small masses.

Specify a specific distribution estimation method (as in LearnDistribution) used to impute missing values.

By default, missing values are imputed by sampling from the learned distribution conditioned on the known values. It is possible to find the most likely imputed values instead. Impute missing values, specifying both the distribution method and the imputation method.