Search

Neural network and machine learning algorithms often have parameters that must be tuned for good performance on a particular task. Leave-one-out cross-validation (LCV) accuracy is often used to measure the fitness of a set of parameter values. However, small changes in parameters often have no effect on LCV accuracy. Many learning algorithms can measure the confidence of a classification decision, but often confidence alone is an inappropriate measure of fitness. This paper proposes a combined measure of cross-validation and confidence (CVC) for obtaining a continuous measure of fitness for sets of parameters in learning algorithms. This paper also proposes the refined instance-based learning algorithm which illustrates the use of CVC in automated parameter tuning. Using CVC provides significant improvement in generalization accuracy on a collection of 31 classification tasks when compared to using LCV.

A relaxation network model that includes higher order weight connections is introduced. To demonstrate its utility, the model is applied to the speech recognition domain. Traditional speech recognition systems typically consider only that context preceding the word to be recognized. However, intuition suggests that considering both preceding context as well as following context should improve recognition accuracy. The work described here tests this hypothesis by applying the higher order relaxation network to consider both precedes and follows context in speech recognition. The results demonstrate both the general utility of the higher order relaxation network as well as its improvement over traditional methods on a speech recognition task.

Constraint satisfaction networks contain nodes that receive weighted evidence from external sources and/or other nodes. A relaxation process allows the activation of nodes to affect neighboring nodes, which in turn can affect their neighbors, allowing information to travel through a network. When doing discrete updates (as in a software implementation of a relaxation network), a goal net or goal activation can be computed in response to the net input into a node, and a relaxation rate can then be used to determine how fast the node moves from its current value to its goal value. An open question was whether or not the relaxation rate is a sensitive parameter. This paper shows that the relaxation rate has almost no effect on how information flows through the network as long as it is small enough to avoid large discrete steps and/or oscillation.