A preprocessing Shapley value-based approach to detect relevant and disparity prone features in machine learning.
Guilherme Dean PelegrinaMiguel CouceiroLeonardo Tomazeli DuartePublished in: FAccT (2024)
Keyphrases
- preprocessing
- machine learning
- feature set
- feature extraction
- feature vectors
- pattern recognition
- learning algorithm
- detection method
- image features
- low level
- data mining
- feature space
- machine learning methods
- decision trees
- artificial intelligence
- multi agent systems
- support vector
- reinforcement learning
- support vector machine
- information extraction
- co occurrence
- computer vision
- detection algorithm
- search engine
- machine learning algorithms
- game theory
- extracted features