Active Learning to Recognize Multiple Types of Plankton

作者： Tong Luo , Kurt Kramer , Dmitry B Goldgof , Lawrence O Hall , Scott Samson

关键词: Class (biology) 、 Image (mathematics) 、 Machine learning 、 Support vector machine 、 Batch processing 、 Least squares support vector machine 、 Semi-supervised learning 、 Active learning (machine learning) 、 Computer science 、 Domain (software engineering) 、 Artificial intelligence 、 Pattern recognition

摘要: This paper presents an active learning method which reduces the labeling effort of domain experts in multi-class classification problems. Active is applied conjunction with support vector machines to recognize underwater zooplankton from higher-resolution, new generation SIPPER II images. Most previous work on only deals two class In this paper, we propose approach "breaking ties" for using one-vs-one a probability approximation. Experimental results indicate that our often requires significantly less labeled images reach given accuracy than least certain test example and random sampling. It can also be batch mode resulting comparable one image at time retraining.