Efficient max-margin multi-label classification with applications to zero-shot learning

By Bharath Hariharan, S. V. N. Vishwanathan, Manik Varma. 2012

68 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Transductive Multi-View Zero-Shot Learning (2015, methodology): “[13] explored the label correlations of auxiliary data via a multi-label max-margin formulation and bet-”
An embarrassingly simple approach to zero-shot learning (2015, background): “Another related approach is proposed in [14], where the authors use the information regarding the correlations between attributes in both training and test instances.”
Multi-label learning with millions of labels: recommending advertiser bid phrases for web pages (2013, methodology): “State-of-the-art discriminative classifiers such as M3L [20] and MetaLabeler [38] would have, for most labels , the daunting task of learning from a data set where the ratio of positive to negative data points is 1 is to 10 million.”
Extreme Multi-label Loss Functions for Recommendation, Tagging, Ranking & Other Missing Label Applications (2016, methodology): “Other baseline algorithms include 1-vs-All [15], LEML [47], WSABIE [44], CPLST [9], CS [17], ML-CSSP [6] and LPSR [45].”
A Review on Multi-Label Learning Algorithms (2014, background): “been some empirical results indicating that simple strategies without sophisticated label correlation modeling would suffice to achieve good performance, such as Binary Relevance with SVM [19] or with mild prior information [65], [66].”
COSTA: Co-Occurrence Statistics for Zero-Shot Classification (2014, methodology): “They have been used to improve image segmentation [11], object detection [3, 22], to reason about what to expect in a scene [4, 14], and for image and attribute classification [8, 15].”
FastXML: a fast, accurate and stable tree-classifier for extreme multi-label learning (2014, methodology): “The hyper-parameters for MLRF, LPSR and M3L were set using fine grained validation on each data set so as to achieve the highest possible prediction accuracy for each method.”
Sparse Local Embeddings for Extreme Multi-label Classification (2015, methodology): “Techniques such as CS [3], CPLST [30], ML-CSSP [7], 1-vs-All [31] could only be trained on the small data sets given standard resources.”