Provable training of a ReLU gate with an iterative non-gradient algorithm.
Sayar KarmakarAnirbit MukherjeePublished in: Neural Networks (2022)
Keyphrases
- learning algorithm
- significant improvement
- k means
- dynamic programming
- optimization algorithm
- similarity measure
- objective function
- detection algorithm
- training phase
- times faster
- worst case
- experimental evaluation
- cost function
- theoretical analysis
- high accuracy
- data sets
- preprocessing
- gradient method
- iterative optimization
- improved algorithm
- recognition algorithm
- approximation algorithms
- supervised learning
- training process
- iterative process
- segmentation algorithm
- itemsets
- linear programming
- probabilistic model
- computational cost
- np hard
- search space
- computational complexity
- search algorithm
- data structure
- neural network