ICLR 2017 accepted papers on Label-efficient and robust learning
24 of the 198 papers accepted at ICLR 2017 (12.1%) are on Label-efficient and robust learning. They fall into 6 topics.
See ICLR 2017 in the interactive map
Submitting to ICLR? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Recognition and long-tail · 16 papers
- A Simple but Tough-to-Beat Baseline for Sentence Embeddings (1,038 citations)
- Visualizing Deep Neural Network Decisions: Prediction Difference Analysis (487 citations)
- Fine-grained Analysis of Sentence Embeddings Using Auxiliary Prediction Tasks (293 citations)
- Offline bilingual word vectors, orthogonal transformations and the inverted softmax (257 citations)
- Generative Models and Model Criticism via Optimized Maximum Mean Discrepancy (163 citations)
- Towards a Neural Statistician (143 citations)
- Learning Features of Music From Scratch (136 citations)
- Neuro-Symbolic Program Synthesis (95 citations)
- Efficient Vector Representation for Documents through Corruption (90 citations)
- Revisiting Classifier Two-Sample Tests (58 citations)
- Predicting Medications from Diagnostic Codes with Recurrent Neural Networks (44 citations)
- Multi-view Recurrent Neural Acoustic Word Embeddings (33 citations)
and 4 more in the interactive map.
Domain adaptation and generalization · 3 papers
- Central Moment Discrepancy (CMD) for Domain-Invariant Representation Learning (213 citations)
- EPOpt: Learning Robust Neural Network Policies Using Model Ensembles (136 citations)
- Attend, Adapt and Transfer: Attentive Deep Architecture for Adaptive Transfer from multiple sources in the same domain (5 citations)
Semi-supervised learning · 2 papers
- Temporal Ensembling for Semi-Supervised Learning (1,481 citations)
- Adversarial Training Methods for Semi-Supervised Text Classification (200 citations)
Self-supervised and contrastive · 1 paper
- Adversarial Feature Learning (689 citations)
Noisy labels · 1 paper
- Training deep neural-networks using a noise adaptation layer (511 citations)
