Evaluation of feature sensitivity to training data inaccuracy in detection of retinal lesions.
Lauri LaaksonenAntti HannukselaEla ClaridgePauli FältMarkku Hauta-KasariHannu UusitaloLasse LensuPublished in: IPTA (2016)
Keyphrases
- training data
- test data
- data sets
- automatic detection
- detection method
- object detection
- false alarm rate
- image analysis
- digital mammograms
- training set
- decision trees
- computer aided detection
- detection algorithm
- target detection
- learning algorithm
- false alarms
- feature values
- fundus images
- machine learning
- computer aided
- supervised learning
- prior knowledge
- feature selection
- sensitivity analysis
- classification models
- evaluation method
- training dataset
- false positives
- support vector machine
- classification accuracy
- lesion segmentation
- adaboost classifier