A Modified General Regression Neural Network (MGRNN) with new, efficient training algorithms as a robust 'black box'-tool for data analysis.
Dirk TomandlAndreas SchoberPublished in: Neural Networks (2001)
Keyphrases
- black box
- computationally efficient
- highly efficient
- general regression neural network
- black boxes
- parameter tuning
- rule extraction
- white box
- computational cost
- computationally expensive
- integration testing
- optimization problems
- training set
- computationally intensive
- data structure
- learning algorithm
- computationally demanding
- neural network
- data sets