TY - GEN
T1 - Exploring the Similarity of Medical Imaging Classification Problems
AU - Cheplygina, Veronika
AU - Moeskops, Pim
AU - Veta, Mitko
AU - Dashtbozorg, Behdad
AU - Pluim, Josien P.W.
PY - 2017
Y1 - 2017
N2 - Supervised learning is ubiquitous in medical image analysis. In this paper we consider the problem of meta-learning – predicting which methods will perform well in an unseen classification problem, given previous experience with other classification problems. We investigate the first step of such an approach: how to quantify the similarity of different classification problems. We characterize datasets sampled from six classification problems by performance ranks of simple classifiers, and define the similarity by the inverse of Euclidean distance in this meta-feature space. We visualize the similarities in a 2D space, where meaningful clusters start to emerge, and show that the proposed representation can be used to classify datasets according to their origin with 89.3% accuracy. These findings, together with the observations of recent trends in machine learning, suggest that meta-learning could be a valuable tool for the medical imaging community.
AB - Supervised learning is ubiquitous in medical image analysis. In this paper we consider the problem of meta-learning – predicting which methods will perform well in an unseen classification problem, given previous experience with other classification problems. We investigate the first step of such an approach: how to quantify the similarity of different classification problems. We characterize datasets sampled from six classification problems by performance ranks of simple classifiers, and define the similarity by the inverse of Euclidean distance in this meta-feature space. We visualize the similarities in a 2D space, where meaningful clusters start to emerge, and show that the proposed representation can be used to classify datasets according to their origin with 89.3% accuracy. These findings, together with the observations of recent trends in machine learning, suggest that meta-learning could be a valuable tool for the medical imaging community.
UR - http://www.scopus.com/inward/record.url?scp=85029820846&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-67534-3_7
DO - 10.1007/978-3-319-67534-3_7
M3 - Conference contribution
AN - SCOPUS:85029820846
SN - 9783319675336
VL - 10552 LNCS
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 59
EP - 66
BT - Intravascular Imaging and Computer Assisted Stenting, and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis - 6th Joint International Workshops, CVII-STENT 2017 and 2nd International Workshop, LABELS 2017 Held in Conjunction with MICCAI 2017, Proceedings
PB - Springer-Verlag
T2 - 6th Joint International Workshops on Computing and Visualization for Intravascular Imaging and Computer Assisted Stenting, CVII-STENT 2017 and 2nd International Workshop on Large-Scale Annotation of Biomedical Data and Expert Label Synthesis, LABELS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017
Y2 - 10 September 2017 through 14 September 2017
ER -