Doubly Adaptive Scaled Algorithm for Machine Learning Using Second-Order Information.
Majid JahaniSergey RusakovZheng ShiPeter RichtárikMichael W. MahoneyMartin TakácPublished in: CoRR (2021)
Keyphrases
- learning algorithm
- machine learning
- high accuracy
- times faster
- optimal solution
- input data
- information extraction
- final result
- detection algorithm
- optimization algorithm
- expectation maximization
- experimental evaluation
- np hard
- preprocessing
- k means
- recognition algorithm
- decision trees
- segmentation algorithm
- web pages
- information sources
- objective function
- knowledge acquisition
- worst case
- improved algorithm
- computational complexity
- computational cost
- domain knowledge
- search space
- information loss
- neural network
- support vector
- convergence rate
- genetic algorithm
- matching algorithm
- theoretical analysis
- linear programming
- simulated annealing
- search algorithm
- higher order
- natural language processing
- significant improvement
- active learning