This module introduces the maximum likelihood estimator. We show how the MLE implements the likelihood principle. Methods for computing th MLE are covered. Properties of the MLE are discussed including asymptotic e ciency and invariance under reparameterization. The maximum likelihood estimator (MLE) is an alternative to the minimum variance unbiased estimator (MVUE). For many estimation proble...

We present a stochastic model of suprathreshold perceptual differences based on difference measurement. We develop a maximum likelihood difference scaling (MLDS) method for estimating its parameters and evaluate the reliability and distributional robustness of the fitting method. We also describe a method for testing whether the difference measurement model is appropriate as a description of hu...

A generalization of the Half Logistic Distribution is developed through exponentiation of its survival function and named the Type II Generalized Half Logistic Distribution (GHLD). The distributional characteristics are presented and estimation of its parameters using maximum likelihood and modified maximum likelihood methods is studied with comparisons. Discrimination between Type II GHLD and ...

Maximum likelihood factor analysis provides an effective method for estimation of factor matrices and a useful test statistic in the likelihood ratio for rejection of overly simple factor models. A reliability coefficient is proposed to indicate quality of representation of interrelations among attributes in a battery by a maximum likelihood factor analysis. Usually, for a large sample of indiv...

This paper presents a novel framework, based on maximum likelihood, for training models to recognise simple spatial-motion events, such as those described by the verbs pick up, put down, push, pull, drop, and throw, and classifying novel observations into previously trained classes. The model that we employ does not presuppose prior recognition or tracking of 3D object pose, shape, or identity....

The problem of estimating the parameters for continuous-time partially observed systems is discussed. New exact lters for obtaining Maximum Likelihood (ML) parameter estimates via the Expectation Maximization algorithm are derived. The methodology exploits relations between incomplete and complete data likelihood and gradient of likelihood functions, which are derived using Girsanov's measure t...

In image matching applications such as tracking and stereo matching, it is common to use the sum-of-squareddi erences (SSD) measure to determine the best match for an image template. However, this measure is sensitive to outliers and is not robust to template variations. We describe a robust measure and eÆcient search strategy for template matching with a binary or greyscale template using a ma...

One popular class of unsupervised algorithms are competitive algorithms. In the traditional view of competition, only one competitor, the winner, adapts for any given case. I propose to view competitive adaptation as attempting to fit a blend of simple probability generators (such as gaussians) to a set of data-points. The maximum likelihood fit of a model of this type suggests a "softer" form ...