Multilabel Classification using Bayesian Compressed Sensing

By Ashish Kapoor, Raajay Viswanathan, Prateek Jain. 2012

139 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 Evaluation & Benchmarks

Citations in the corpus

Elucidating image-to-set prediction: An analysis of models, losses and datasets (2019, background): “Multi-label classification has been a long lasting problem in computer vision [72, 21, 73, 67] tackled by a wide variety of methods, such as, decomposing the problem into independent single-label classification problems [44, 72], exploiting label co-occurrences [1, 56, 34], introducing priors such a”
Structured Prediction Energy Networks (2015, methodology): “Motivated by recent compressed sensing approaches to multi-label classiﬁcation (Hsu et al., 2009; Kapoor et al., 2012), we further assume that the ﬁrst layer of the network performs a small set of linear projections of the prediction variables.”
FastXML: a fast, accurate and stable tree-classifier for extreme multi-label learning (2014, methodology): “Methods mainly differ in the choice of compression and decompression techniques such as compressed sensing [19, 22], Bloom filters [12], SVD [29], landmark labels [4, 8], output codes [38], etc.”
Multi-label learning with millions of labels: recommending advertiser bid phrases for web pages (2013, background): “Almost all the other work on large category multi-label classification deals with thousands of labels [21,24,41,42] and it is not immediately clear that these techniques will scale to millions of labels and beyond.”
Extreme Multi-label Loss Functions for Recommendation, Tagging, Ranking & Other Missing Label Applications (2016, background): “As such, the Hamming loss [4, 6, 9, 10, 13, 18, 21, 39, 42, 47] continues to be one of the most popular losses for extreme multi-label learning along with precision [3, 5, 17, 20, 21, 33, 44, 45] and the F-measure [4, 10, 14, 18, 19, 21, 39, 49].”
CU-Net: Component Unmixing Network for Textile Fiber Identification (2019, methodology): “These approaches usually exploit distinctive properties of multilabel classification, such as label sparsity (Song et al. 2015; Hsu et al. 2009; Kapoor et al. 2012; Yang et al. 2014, 2018; Zhou and Tao 2012; Zhang et al. 2015), label dependency (Shu et al. 2015; Alessandro et al. 2013; Liu and…”
Towards Scalable and Reliable Capsule Networks for Challenging NLP Applications (2019, methodology): “Label embedding models (Balasubramanian and Lebanon, 2012; Chen and Lin, 2012; Cisse et al., 2013; Bi and Kwok, 2013; Ferng and Lin, 2011; Hsu et al., 2009; Ji et al., 2008; Kapoor et al., 2012; Lewis et al., 2004; Yu et al., 2014a) address the data sparsity issue with two steps: compression and dec”
Deep Learning for Extreme Multi-label Text Classification (2017, methodology): “Many variants of the target embedding methods have been proposed [3, 6, 9, 10, 14, 19, 20, 24, 40, 46, 50].”
Large-scale Multi-label Learning with Missing Labels (2013, methodology): “…four standard datasets (bibtex, deli-cious, eurlex, and nus-wide), two datasets with d L (autofood and compphys), and a very large scale CS Hsu et al. (2009) and PLST Tai & Lin (2012) are not included as they are shown to be suboptimal to CPLST and BCS in Chen & Lin (2012); Kapoor et al. (2012).”
Exploiting sparsity and co-occurrence structure for action unit recognition (2015, methodology): “Further, recent advances in compressed sensing [13], [16] have shown how much computational efficiency such sparsity provides without compromising the quality of the results.”
Reading recognition of pointer meter based on pattern recognition and dynamic three-points on a line (2017, methodology): “All of the existing methods of recognizing pointer meter reading based on computer vision are inclined to select a certain type of instrument as a modelling object, among which Table Look-up method and Static Three Points Make a Line(STMLP) are typical examples [5].”
Multi-label Classification via Feature-aware Implicit Label Space Encoding (2014, methodology): “As revealed by Kapoor et al. (2012), the large vocabularies force many existing effective multi-label classification methods (Boutell et al., 2004; Evgeniou et al., 2005; Tsochantaridis et al., 2006; Tsoumakas & Katakis, 2007; Argyriou et al., 2008; Hariharan et al., 2010; Tsoumakas et al., 2010;…”
Improving Pairwise Ranking for Multi-label Image Classification (2017, background): “label dependency [1, 25], label sparsity [10, 12, 27], and label noise [33, 39].”
Multi-label Learning with Missing Labels Using Mixed Dependency Graphs (2018, methodology): “Third, missing labels are treated as latent variables in probabilistic models, including the model based on Bayesian networks (Kapoor et al. 2012; Vasisht et al. 2014) and conditional restricted Boltz-mann machines (CR-BM).”