Galaxy model parameters using numerical maximum likelihood estimation

We discuss a numerical algorithm based on maximum likelihood for estimating parameters for models of the Galaxy. We use simultaneously all the information available in a catalog of stellar data in a global optimization, to derive unbiased estimates of intrinsic stellar properties, such as luminosity and velocity dispersion. The likelihood function is defined in the observed domain using quantities such as photoelectric photometry, line‐of‐sight velocity, proper motion, trigonometric parallax, and metallicity. Individual stars included in the statistical analysis can have different amounts of information available. This method includes an explicit treatment of observational errors, can identify outliers objectively and allows use of stellar data with relatively large errors. It can self‐consistently detect and correct for systematic deviations in the observations, such as zero point residuals or underestimated errors.