Semi-supervised Ranking Pursuit
This work addresses ranking and preference learning problems, particularly benefiting scenarios with limited labeled data, though it appears incremental as it builds on existing kernel matching pursuit methods.
The paper tackles the problem of sparse preference learning and ranking by proposing a novel algorithm that approximates the utility function using a weighted sum of basis functions with squared loss on pairs, generalizing kernel matching pursuit. The result shows that the algorithm outperforms state-of-the-art methods when using unlabeled data and performs comparably in supervised settings while providing sparser solutions.
We propose a novel sparse preference learning/ranking algorithm. Our algorithm approximates the true utility function by a weighted sum of basis functions using the squared loss on pairs of data points, and is a generalization of the kernel matching pursuit method. It can operate both in a supervised and a semi-supervised setting and allows efficient search for multiple, near-optimal solutions. Furthermore, we describe the extension of the algorithm suitable for combined ranking and regression tasks. In our experiments we demonstrate that the proposed algorithm outperforms several state-of-the-art learning methods when taking into account unlabeled data and performs comparably in a supervised learning scenario, while providing sparser solutions.