AAAI 2021 papers
Dynamic systems
- On the Verification of Neural ODEs with Stochastic Guarantees
- Forecasting Reservoir Inflow via Recurrent Neural ODEs
- The LOB Recreation Model: Predicting the Limit Order Book from TAQ History Using an Ordinary Differential Equation Recurrent Neural Network
- ECG ODE-GAN: Learning Ordinary Differential Equations of ECG Dynamics via Generative Adversarial Learning
Vid-ODE: Continuous-Time Video Generation with Neural Ordinary Differential Equation
A Hybrid Stochastic Gradient Hamiltonian Monte Carlo Method
Generative models
- Flow-Based Generative Models for Learning Manifold to Manifold Mappings
- OT-Flow: Fast and Accurate Continuous Normalizing Flows via Optimal Transport
- MolGrow: A Graph Normalizing Flow for Hierarchical Molecular Generation
- Accelerating Continuous Normalizing Flow with Trajectory Polynomial Regularization
Time Series
- Deep Switching Auto-Regressive Factorization: Application to Time Series Forecasting
- Dynamic Gaussian Mixture Based Deep Generative Model for Robust Forecasting on Sparse Multivariate Time Series
- Second Order Techniques for Learning Time-Series with Structural Breaks
- Correlative Channel-Aware Fusion for Multi-View Time Series Classification
- Learnable Dynamic Temporal Pooling for Time Series Classification
- Learning Representations for Incomplete Time Series Clustering
- Temporal Latent Autoencoder: A Method for Probabilistic Multivariate Time Series Forecasting
- Continuous-Time Attention for Sequential Learning
- Graph Neural Network-Based Anomaly Detection in Multivariate Time Series
- ShapeNet: A Shapelet-Neural Network Approach for Multivariate Time Series Classification
- Time Series Anomaly Detection with Multiresolution Ensemble Decoding
- Joint-Label Learning by Dual Augmentation for Time Series Classification
- Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting
- Generative Semi-Supervised Learning for Multivariate Time Series Imputation
- Outlier Impact Characterization for Time Series Data
- Meta-Learning Framework with Applications to Zero-Shot Time-Series Forecasting
GNN
- Contrastive and Generative Graph Convolutional Networks for Graph-Based SemiSupervised Learning
- Overcoming Catastrophic Forgetting in Graph Neural Networks
- Isolation Graph Kernel
- Semi-Supervised Node Classification on Graphs: Markov Random Fields vs. Graph Neural Networks
- Hyperbolic Variational Graph Neural Network for Modeling Dynamic Graphs
- MolGrow: A Graph Normalizing Flow for Hierarchical Molecular Generation
- Heterogeneous Graph Structure Learning for Graph Neural Networks
- Graph Neural Network-Based Anomaly Detection in Multivariate Time Series
- Deep Graph Spectral Evolution Networks for Graph Topological Evolution
- Scalable Graph Networks for Particle Simulations
- Synchronous Dynamical Systems on Directed Acyclic Graphs: Complexity and Algorithms
- Fitting the Search Space of Weight-Sharing NAS with Graph Convolutional Networks
- Contrastive Self-Supervised Learning for Graph Classification
- Computationally Tractable Riemannian Manifolds for Graph Embeddings\
- Graph Neural Networks with Heterophily
- Learning Graph Neural Networks with Approximate Gradient Descent
- Probabilistic Dependency Graphs
- Scalable and Explainable 1-Bit Matrix Completion via Graph Signal Learning
- GraphMix: Improved Training of GNNs for Semi-Supervised Learning
- Power up! Robust Graph Convolutional Network via Graph Powering
- Identity-Aware Graph Neural Networks
- Beyond Low-Frequency Information in Graph Convolutional Networks
- Rethinking Graph Regularization for Graph Neural Networks
Optimiser
ADAHESSIAN: An Adaptive Second Order Optimizer for Machine Learning
