real-world data
Data collected from actual scenarios as opposed to simulated environments. In AI, real-world data is critical for training and evaluating models to ensure they perform satisfactorily in practical applications.
- Adversarial generalization of unfolding (model-based) networks
- ConTextTab: A Semantics-Aware Tabular In-Context Learner
- Coupling Generative Modeling and an Autoencoder with the Causal Bridge
- Distributionally Robust Feature Selection
- DoseSurv: Predicting Personalized Survival Outcomes under Continuous-Valued Treatments
- Forecasting in Offline Reinforcement Learning for Non-stationary Environments
- Gene Regulatory Network Inference in the Presence of Selection Bias and Latent Confounders
- Generalized Top-k Mallows Model for Ranked Choices
- Learning-Augmented Online Bipartite Fractional Matching
- Network two-sample test for block models
- Neural Mutual Information Estimation with Vector Copulas
- Optimal Neural Compressors for the Rate-Distortion-Perception Tradeoff
- Quantifying Uncertainty in the Presence of Distribution Shifts
- When Additive Noise Meets Unobserved Mediators: Bivariate Denoising Diffusion for Causal Discovery