Stochasticity helps to navigate rough landscapes: comparing gradient-descent-based algorithms in the phase retrieval problem.
Francesca MignaccoPierfrancesco UrbaniLenka ZdeborováPublished in: Mach. Learn. Sci. Technol. (2021)
Keyphrases
- orders of magnitude
- learning algorithm
- data structure
- machine learning
- information retrieval
- computationally efficient
- worst case
- optimization problems
- preprocessing phase
- neural network
- theoretical analysis
- information retrieval systems
- image database
- rough sets
- computational cost
- cost function
- significant improvement
- retrieval systems
- computational efficiency
- combinatorial optimization
- text retrieval
- image retrieval