Prediction of learning process of human-machine interface with intermissions through a neural network

In order to adapt a human-machine interface to individual user's learning condition, while enabling the user to easily use the interface, the individual learning process should be studied. After a long term intermission in operating a machine, the efficiency of the machine operation may worsen because the intermission weakens the learning results. In this research a hierarchical neural network with an intermediate layer has been developed in order to forecast the user's learning capability after the recommencement of the operation, based on the data gathered in previous operations. The number of units in the intermediate layer was determined by cross validating the data of experiments.