Solving the linear interval tolerance problem for weight initialization of neural networks.
Stavros P. AdamDimitrios A. KarrasGeorge D. MagoulasMichael N. VrahatisPublished in: Neural Networks (2014)
Keyphrases
- neural network
- global exponential stability
- pattern recognition
- convex quadratic programming
- inverse problems in image processing
- fuzzy logic
- back propagation
- quadratic programming
- neural nets
- genetic algorithm
- weight update
- initial conditions
- weight vector
- linear systems
- feed forward
- neural network model
- self organizing maps
- multi layer perceptron
- linear model
- algebraic equations
- highly non linear
- combinatorial optimization
- learning algorithm