Probing Classifiers are Unreliable for Concept Removal and Detection.
Abhinav KumarChenhao TanAmit SharmaPublished in: NeurIPS (2022)
Keyphrases
- detection method
- training data
- detection rate
- object detection
- detection algorithm
- decision trees
- classification systems
- test set
- machine learning algorithms
- false positives
- training set
- support vector
- naive bayes
- training samples
- face detection
- computer vision
- feature set
- svm classifier
- object recognition
- concept learning
- false alarms
- detection accuracy
- linear classifiers
- concept detection
- support vector machine classifier
- adaboost classifier