Overfitting, robustness, and malicious algorithms: A study of potential causes of privacy risk in machine learning.
Samuel YeomIrene GiacomelliAlan MenagedMatt FredriksonSomesh JhaPublished in: J. Comput. Secur. (2020)
Keyphrases
- machine learning
- learning algorithm
- computationally efficient
- machine learning approaches
- decision trees
- machine learning methods
- pattern recognition
- machine learning algorithms
- computational cost
- theoretical analysis
- optimization problems
- learning systems
- empirical studies
- data sets
- support vector machine
- significant improvement
- training data
- neural network
- text classification
- computational intelligence
- upper bound
- reinforcement learning
- computational efficiency
- times faster
- combinatorial optimization
- statistical methods