TY - JOUR
T1 - Cost-Sensitive Learning of Deep Feature Representations From Imbalanced Data
AU - Khan, Salman
AU - HAYAT, Munawar
AU - Bennamoun, Mohammed
AU - Sohel, Ferdous
AU - Togneri, Roberto
PY - 2018/8
Y1 - 2018/8
N2 - Class imbalance is a common problem in the case of real-world object detection and classification tasks. Data of some classes are abundant, making them an overrepresented majority, and data of other classes are scarce, making them an underrepresented minority. This imbalance makes it challenging for a classifier to appropriately learn the discriminating boundaries of the majority and minority classes. In this paper, we propose a cost-sensitive (CoSen) deep neural network, which can automatically learn robust feature representations for both the majority and minority classes. During training, our learning procedure jointly optimizes the class-dependent costs and the neural network parameters. The proposed approach is applicable to both binary and multiclass problems without any modification. Moreover, as opposed to data-level approaches, we do not alter the original data distribution, which results in a lower computational cost during the training process. We report the results of our experiments on six major image classification data sets and show that the proposed approach significantly outperforms the baseline algorithms. Comparisons with popular data sampling techniques and CoSen classifiers demonstrate the superior performance of our proposed method.
AB - Class imbalance is a common problem in the case of real-world object detection and classification tasks. Data of some classes are abundant, making them an overrepresented majority, and data of other classes are scarce, making them an underrepresented minority. This imbalance makes it challenging for a classifier to appropriately learn the discriminating boundaries of the majority and minority classes. In this paper, we propose a cost-sensitive (CoSen) deep neural network, which can automatically learn robust feature representations for both the majority and minority classes. During training, our learning procedure jointly optimizes the class-dependent costs and the neural network parameters. The proposed approach is applicable to both binary and multiclass problems without any modification. Moreover, as opposed to data-level approaches, we do not alter the original data distribution, which results in a lower computational cost during the training process. We report the results of our experiments on six major image classification data sets and show that the proposed approach significantly outperforms the baseline algorithms. Comparisons with popular data sampling techniques and CoSen classifiers demonstrate the superior performance of our proposed method.
KW - Convolutional neural networks (CNNs)
KW - cost-sensitive (CoSen) learning
KW - data imbalance
KW - loss functions
UR - http://www.scopus.com/inward/record.url?scp=85028455314&partnerID=8YFLogxK
UR - http://www.mendeley.com/research/costsensitive-learning-deep-feature-representations-imbalanced-data
U2 - 10.1109/TNNLS.2017.2732482
DO - 10.1109/TNNLS.2017.2732482
M3 - Article
SN - 2162-2388
VL - 29
SP - 3573
EP - 3587
JO - IEEE Transactions on Neural Networks and Learning Systems
JF - IEEE Transactions on Neural Networks and Learning Systems
IS - 8
ER -