paper-with-me

홈 › Papers

SoDeep: a Sorting Deep net to learn ranking loss surrogates

2019-04-08 · CVPR 2019 6 · Martin Engilberge, Louis Chevallier, Patrick Pérez, Matthieu Cord

Several tasks in machine learning are evaluated using non-differentiable metrics such as mean average precision or Spearman correlation. However, their non-differentiability prevents from using them as objective functions in a learning framework. Surrogate and relaxation methods exist but tend to be specific to a given metric. In the present work, we introduce a new method to learn approximations of such non-differentiable objective functions. Our approach is based on a deep architecture that approximates the sorting of arbitrary sets of scores. It is trained virtually for free using synthetic data. This sorting deep (SoDeep) net can then be combined in a plug-and-play manner with existing deep architectures. We demonstrate the interest of our approach in three different tasks that require ranking: Cross-modal text-image retrieval, multi-label image classification and visual memorability ranking. Our approach yields very competitive results on these three tasks, which validates the merit and the flexibility of SoDeep as a proxy for sorting operation in ranking-based losses.

📄 PDF Abstract BibTeX arXiv:1904.04272

Code (1)

technicolor-research/sodeep pytorch

Tasks

image-classificationImage ClassificationImage RetrievalMulti-Label Image ClassificationRetrieval

Similar Papers 제목 키워드 기반

PiRank: Scalable Learning To Rank via Differentiable Sorting

2020-12-12 · NeurIPS 2021 12 · Robin Swezey, Aditya Grover, Bruno Charron, Stefano Ermon

A key challenge with machine learning approaches for ranking is the gap between the performance metrics of interest and the surrogate loss functions that can be optimized with gradient-based methods. This gap arises beca…

Learning-To-Rank

An Embedding Framework for Consistent Polyhedral Surrogates

2019-07-17 · NeurIPS 2019 12 · Jessie Finocchiaro, Rafael Frongillo, Bo Waggoner

We formalize and study the natural approach of designing convex surrogate loss functions via embeddings, for problems such as classification, ranking, or structured prediction. In this approach, one embeds each of the fi…

Structured Prediction

An Embedding Framework for the Design and Analysis of Consistent Polyhedral Surrogates

2022-06-29 · Jessie Finocchiaro, Rafael M. Frongillo, Bo Waggoner

We formalize and study the natural approach of designing convex surrogate loss functions via embeddings, for problems such as classification, ranking, or structured prediction. In this approach, one embeds each of the fi…

Structured Prediction

Convex Calibrated Surrogates for Low-Rank Loss Matrices with Applications to Subset Ranking Losses

2013-12-01 · NeurIPS 2013 12 · Harish G. Ramaswamy, Shivani Agarwal, Ambuj Tewari

The design of convex, calibrated surrogate losses, whose minimization entails consistency with respect to a desired target loss, is an important concept to have emerged in the theory of machine learning in recent years. …

Perceptron-like Algorithms and Generalization Bounds for Learning to Rank

2014-05-03 · Sougata Chaudhuri, Ambuj Tewari

Learning to rank is a supervised learning problem where the output space is the space of rankings but the supervision space is the space of relevance scores. We make theoretical contributions to the learning to rank prob…

Generalization BoundsLearning-To-RankStructured Prediction