Explanation-based Learning with Feedforward Neural Networks
This thesis primarily attempts to incorporate an argument loss in feedforward networks, in order maximize the contributions of selected input attributes to the respective prediction.
true" ? copyright : '' }