Dataset Properties and Degradation of Machine Learning Accuracy with an Anonymized Training Dataset.
Wakana MaedaToshiya ShimizuTakeru FukuokaIkuya MorikawaPublished in: CANDAR (Workshops) (2020)
Keyphrases
- training dataset
- machine learning
- training data
- high accuracy
- training samples
- training set
- data samples
- class labels
- classification accuracy
- learning algorithm
- feature selection
- machine learning algorithms
- support vectors
- supervised learning
- information extraction
- support vector machine
- computer vision
- imbalanced datasets
- prediction accuracy
- text classification
- natural language processing