Super-sparse Learning In Similarity Spaces

Abstract

In several applications, input samples are more naturally represented in terms of similarities between each other, rather than in terms of feature vectors. In these settings, machine-learning algorithms can become very computationally demanding, as they may require matching the test samples against a very large set of reference prototypes. To mitigate this issue, different approaches have been developed to reduce the number of required reference prototypes. Current reduction approaches select a small subset of representative prototypes in the space induced by the similarity measure, and then separately train the classification function on the reduced subset. However, decoupling these two steps may not allow reducing the number of prototypes effectively without compromising accuracy. We overcome this limitation by jointly learning the classification function along with an optimal set of virtual prototypes, whose number can be either fixed a priori or optimized according to application-s

Stats

• citations4
• S2 citations—
• github stars0
• HF likes0
• heat score5.24
• arxiv keydemontis2017super

Related papers

• Sparse Online Relative Similarity Learning (2021)2.26
• A New Similarity Space Tailored For Supervised Deep Metric Learning (2020)3.58
• Similarity Learning Via Kernel Preserving Embedding (2019)10.35
• Escaping The Curse Of Dimensionality In Similarity Learning: Efficient Frank-wolfe Algorithm And Generalization Bounds (2018)9.41
• Visual Space Optimization For Zero-shot Learning (2019)0.00
• Discriminative Learning Of Similarity And Group Equivariant Representations (2018)0.00
• Pairwise Similarity Learning Is Simple (2023)7.50
• Efficient Similarity-preserving Unsupervised Learning Using Modular Sparse Distributed Codes And Novelty-contingent Noise (2020)0.00