next up previous contents
Next: 7.2.7 Anticipation performance Up: 7.2 Methods Previous: 7.2.5 Forward model: Abstract

7.2.6 Performance outside the training domain

Both the abstract RNN and the MLP were only trained on data points whose sensory components were restricted to a two-dimensional manifold. The free parameters are the robot's distance to the center of the circle and the robot's orientation. Since it is not clear how a network reacts to points slightly outside its training domain, we have a look at the effect of a small change in the input of the forward model.

Let $ \bf f$($ \bf s$) be the transformation the network does on the sensory input. To each sensory input $ \bf s$ from the test set, in ten trials, a divergence $ \bf e$ was added. This divergence was distributed randomly and extended uniformly into the ten-dimensional sensory subspace. The magnitude of $ \bf e$ ranged between 0.0 and 1.0 pixels. The computation of $ \bf f$($ \bf s$) also requires a pair of velocities; in each trial, they were chosen randomly from the interval [-60; 60]. The results are in section 7.3.3.


next up previous contents
Next: 7.2.7 Anticipation performance Up: 7.2 Methods Previous: 7.2.5 Forward model: Abstract
Heiko Hoffmann
2005-03-22