Learning Long-Term Dependencies with Gradient Descent Is Difficult

Venue

Publication Year

Keywords

Computer networks,Cost function,Delay effects,Discrete transforms,Displays,efficient learning,gradient descent,input/output sequence mapping,Intelligent networks,learning (artificial intelligence),long-term dependencies,Neural networks,Neurofeedback,numerical analysis,prediction problems,Production,production problems,recognition,recurrent neural nets,recurrent neural network training,Recurrent neural networks,temporal contingencies

Identifiers

Authors

  • Y. Bengio
  • P. Simard
  • P. Frasconi

Source Materials