arXiv 2018

47,976 papers
In 2018, arXiv hosted 2,294 papers, with a significant focus on advancing optimization techniques for neural networks and large-scale systems. Key research directions included scalable second-order optimization, local curvature adaptation, and decentralized optimization strategies that account for topology and spectral gaps. Scholars such as Wotao Yin, Michael Jordan, and Jianqing Fan contributed to these developments, alongside work on robust optimization using inexact oracles and proximal operators. There was also a notable emphasis on statistical learning theory, particularly concerning dependent dynamical systems and convergence to second-order stationarity in non-convex optimization. Research also prioritized scaling Bayesian inference through stochastic gradient MCMC and addressing computationally expensive simulators. Gaussian processes were explored with a focus on overcoming cubic complexity limitations. Applied research spanned diverse fields, including wireless spectrum anomaly detection, renewable energy forecasting, and the study of eye movement influences. Additionally, papers examined the intersection of economics and engineering, such as token economics in energy systems and the relationship between Bayes A-optimal designs and the Group Lasso. Other contributions investigated signal waveform recovery and diffusion coefficients in multi-component glass-forming liquids.

Top Research Directions

Research directions most represented by arXiv 2018 papers.

Generalizing from Synthetic Data to the Real World 181
Preserving Tiny Features Against Spatial Resolution Degradation 152
Sample Efficient Reinforcement Learning 127
Mitigating Input Perturbation Sensitivity in Deep Networks 126
Scaling Gaussian Processes Beyond Cubic Complexity 122
Dynamic Preference Modeling in Interactive Systems 102
Self-Explainable Neural Networks without Performance Loss 100
Clean Accuracy Degradation in Machine Learning Defenses 99
Compressing Deep Networks for Edge Hardware Constraints 98
Bypassing the Computational Bottleneck of Adversarial Training 96
The Empirical Validation Gap 93
Parallelizing Recurrent Models via Linear State Transitions 92
Adversarial Training for Domain Invariant Representations 87
Overcoming Volumetric Context Loss in Medical Vision Models 86
Overcoming the Convexity Constraint in Modern Optimization 86

Top Authors

AuthorPapers
Yoshua Bengio68
Sergey Levine66
Philip S. Yu55
Xiaojiang Du49
Wei Liu44
Sven Behnke43
Dacheng Tao42
Nassir Navab41
Eric P. Xing38
Mohsen Guizani38
Walid Saad38
Mohamed-Slim Alouini37
Graham Neubig36
Luc Van Gool34
Dusit Niyato33
Alan Yuille32
Xu Sun32
Bernhard Schölkopf31
Ming-Hsuan Yang31
Chunhua Shen30

Impactful Papers

Papers from arXiv 2018 that shaped research clusters.

Foundational

Influential

Papers

Title
3D RoI-aware U-Net for Accurate and Efficient Colorectal Tumor Segmentation
3D Semi-Supervised Learning with Uncertainty-Aware Multi-View Co-Training
3D Shape Perception from Monocular Vision, Touch, and Shape Priors
3D Shape Reconstruction from a Single 2D Image via 2D-3D Self-Consistency
3D-SIS: 3D Semantic Instance Segmentation of RGB-D Scans
3D Snapshot Microscopy of Extended Objects
3DSRnet: Video Super-resolution using 3D Convolutional Neural Networks
3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data
3D super-resolved multi-angle TIRF via polarization modulation
3D Surface Reconstruction by Pointillism
3D Terrain Segmentation in the SWIR Spectrum
3DTI-Net: Learn Inner Transform Invariant 3D Geometry Features using Dynamic GCN
3d TQFTs from Argyres-Douglas theories
3D Traffic Simulation for Autonomous Vehicles in Unity and Python
3-wise Independent Random Walks can be Slightly Unbounded
4D Human Body Correspondences from Panoramic Depth Maps
4D Temporally Coherent Light-field Video
50 Years of Test (Un)fairness: Lessons for Machine Learning
5G Wireless Network Slicing for eMBB, URLLC, and mMTC: A Communication-Theoretic View
82 Treebanks, 34 Models: Universal Dependency Parsing with Multi-Treebank Models
8T SRAM Cell as a Multi-bit Dot Product Engine for Beyond von-Neumann Computing
99\% Revenue via Enhanced Competition

« prev Page 4/960 next »