Maximizing Subset Accuracy with Recurrent Neural Networks in Multi-label Classification

By Jinseok Nam, Eneldo Loza Mencía, Hyunwoo J. Kim, Johannes Fürnkranz. 2017

184 citations recorded by the source corpus.

Subfields: Model Architectures, Evaluation & Benchmarks

How researchers see this paper

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

In Model Architectures

In Evaluation & Benchmarks

Citations in the corpus

Elucidating image-to-set prediction: An analysis of models, losses and datasets (2019, neutral): “In order to explicitly exploit label co-occurrences, researchers have resorted to modeling the joint probability distribution of labels [59] or decomposing the joint distribution into conditionals [11, 61, 45, 28, 32, 41, 46].”
Seq2Slate: Re-ranking and Slate Optimization with RNNs (2018, methodology): “Sensitivity to input order Previous work suggests that the performance of seq2seq models is often sensitive to the order in which the input is processed [Vinyals et al., 2016, Nam et al., 2017, Ai et al., 2018a].”
Asymmetric Loss For Multi-Label Classification (2020, methodology): “Other approaches are based on modeling image parts and attentional regions [37, 13, 33, 36], as well as using recurrent neural networks [25, 30].”
Two-Stage Label Embedding via Neural Factorization Machine for Multi-Label Classification (2019, background): “Several state-of-the-art techniques (Zhang and Zhou 2006; Wang et al. 2016; Yeh et al. 2017; Nam et al. 2017) exploit deep neural network (DNN) for MLC.”
Inverse Cooking: Recipe Generation From Food Images (2018, methodology): “Significant effort has been devoted in the literature to leverage deep neural networks for multi-label classification, by designing models [49, 8, 56, 37, 53] and studying loss functions [12] well suited for this task.”
Deep Structured Prediction with Nonlinear Output Transformations (2018, background): “It has achieved impressive results in machine translation [51, 28], computer vision [38], and multi-label classification [36].”
Loss Functions for Multiset Prediction (2017, methodology): “set prediction) was posed as sequence prediction with RNNs [24, 14], improving upon methods that do not model conditional label dependencies.”
The Emerging Trends of Multi-Label Learning (2020, methodology): “Recently advanced deep learning architectures [37], [38], [39], [40] for MLC problems are studied.”
Deep Streaming Label Learning (2020, background): “Note that SLEEC, SML, and SLL only focus on ranking results; if adopting a naive thresholding policy (e.g., a threshold of 0.5), they underperform compared to baselines, i.e., BR (Nam et al., 2017).”
Neural Message Passing for Multi-Label Classification (2019, methodology): “The first group, probabilistic classifier chain (PCC) models, formulate the joint label dependencies using the chain rule and perform MLC in a sequential prediction manner [39, 55, 36].”
General Multi-label Image Classification with Transformers (2020, background): “(1) Conditional Prediction: The ﬁrst type, autoregressive models [11, 39, 48, 37] estimate the true joint probability of output labels given the input by using the chain rule, predicting one label at a time.”
AttentionXML: Label Tree-based Attention-Aware Deep Model for High-Performance Extreme Multi-Label Text Classification (2018, methodology): “Note that both MLC2Seq and XML-CNN are unable to deal with datasets with millions of labels.”