Principled Weight Initialisation for Input-Convex Neural Networks.
Pieter-Jan HoedtGünter KlambauerPublished in: CoRR (2023)
Keyphrases
- neural network
- pattern recognition
- desired output
- genetic algorithm
- neural nets
- neural network model
- machine learning
- hidden units
- nonlinear functions
- learning rules
- convex optimization
- convex hull
- input data
- expert systems
- decision trees
- fully automatic
- multilayer perceptron
- markov random field
- input variables
- rbf network
- computational intelligence
- user input
- artificial neural networks
- weight update