This package contains MATLAB implementation of saccade model, an algorithm to recognize a good single saccade from a set of points. This implementation was later on converted and developed further in Python. The python version and further details of the functioning of the model are available at github.com/infant-cognition-tampere/saccademodel-py.
The algorithm is two-part. The file saccadeMLE.m defines the model and computes mean square error (MSE) of model's fit for a single set of model parameters. The file saccadeEM.m defines a procedure to iteratively find the set of model parameters that yield the smallest MSE. The files mletest.m and emtest.m provide usage examples.
Akseli Palén, akseli.palen@gmail.com
Infant Cognition Laboratory at University of Tampere
MIT License