Positive semidefinite support vector regression metric learning
This work addresses a specific weakness in metric learning for scenarios like multi-label learning, but it is incremental as it builds directly on the existing RAML framework.
The paper tackled the limitation of the RAML framework in learning positive semidefinite distance metrics by proposing two new methods, which achieved favorable performance in experiments on single-label classification, multi-label classification, and label distribution learning.
Most existing metric learning methods focus on learning a similarity or distance measure relying on similar and dissimilar relations between sample pairs. However, pairs of samples cannot be simply identified as similar or dissimilar in many real-world applications, e.g., multi-label learning, label distribution learning. To this end, relation alignment metric learning (RAML) framework is proposed to handle the metric learning problem in those scenarios. But RAML framework uses SVR solvers for optimization. It can't learn positive semidefinite distance metric which is necessary in metric learning. In this paper, we propose two methds to overcame the weakness. Further, We carry out several experiments on the single-label classification, multi-label classification, label distribution learning to demonstrate the new methods achieves favorable performance against RAML framework.