arXiv research

Model predicts counterfactuals under domain shift and inaccessible variables.

Improved learning bounds for corrupted data using thresholded gradient descent.

New algorithm reduces runtime for robust sparse mean estimation.

New algorithm learns Bayesian networks robustly in nearly-linear time.

COLUMBUS discovers new features to improve domain generalization.

Framework insures AI actions with reserve capital, preventing loss.

A new algorithm reduces online exp-concave optimization runtime.

Evolutionary algorithm finds optimal pixel perturbations to improve neural network generalization.

We study the problem of estimating the covariance matrix of a high-dimensional distribution when a small constant fraction of the samples can be arbitrarily corrupted. Recent work gave the first polynomial time algorithms for this problem with near-optimal error guarantees for several natural structured distributions. …

Achieving robustness to distributional shift is a longstanding and challenging goal of computer vision. Data augmentation is a commonly used approach for improving robustness, however robustness gains are typically not uniform across corruption types. Indeed increasing performance in the presence of random noise is oft…

RILA learns HQMMs robustly against adversarial corruption.

Paper tackles noise-robust domain adaptation in noisy environments.

SSMBA generates synthetic data to improve robustness in natural language tasks.

Improves tabular data augmentation for contrastive learning.

Improved robust regression algorithms with faster runtime and better estimation rates.

Improves Gaussian process factor models for multi-population recordings.

For well over a quarter century, detection systems have been driven by models learned from input features collected from real or simulated environments. An artifact (e.g., network event, potential malware sample, suspicious email) is deemed malicious or non-malicious based on its similarity to the learned model at runt…

New method generates clean data from corrupted observations.

New algorithm estimates robust Gaussian covariance in nearly matrix multiplication time.

New algorithm clusters data with almost-linear time, robust to corruption.

New robust loss functions improve matrix completion accuracy.

Simple method improves uncertainty estimation for distribution shifts.

Superpixel-mix enhances reliability in semantic segmentation.

We study the problem of robust time series analysis under the standard auto-regressive (AR) time series model in the presence of arbitrary outliers. We devise an efficient hard thresholding based algorithm which can obtain a consistent estimate of the optimal AR model despite a large fraction of the time series points …

New method for robust linear regression in nearly linear time.

NoisyMix boosts model robustness to common corruptions.

ZICO learns DAGs from zero-inflated count data efficiently.

New method decomposes corrupted data matrices into sparse and low-rank components.

Study proposes worst+gap measure for better DG evaluation.

In the past few years, Generative Adversarial Networks (GANs) have dramatically advanced our ability to represent and parameterize high-dimensional, non-linear image manifolds. As a result, they have been widely adopted across a variety of applications, ranging from challenging inverse problems like image completion, t…

CANDI solves the gap between continuous and discrete diffusion models for text generation.

PClean automates Bayesian data cleaning for specific datasets.

PDSketch enables flexible robot planning by learning from domain structures.

Develops a comprehensive theory of corruption in supervised learning.

Quantifying the value of data is a fundamental problem in machine learning. Data valuation has multiple important use cases: (1) building insights about the learning task, (2) domain adaptation, (3) corrupted sample discovery, and (4) robust learning. To adaptively learn data values jointly with the target task predict…

The success of generative modeling in continuous domains has led to a surge of interest in generating discrete data such as molecules, source code, and graphs. However, construction histories for these discrete objects are typically not unique and so generative models must reason about intractably large spaces in order…

For using neural networks in safety critical domains, it is important to know if a decision made by a neural network is supported by prior similarities in training. We propose runtime neuron activation pattern monitoring - after the standard training process, one creates a monitor by feeding the training data to the ne…

The problem of domain generalization is to take knowledge acquired from a number of related domains where training data is available, and to then successfully apply it to previously unseen domains. We propose a new feature learning algorithm, Multi-Task Autoencoder (MTAE), that provides good generalization performance …

Improved covariate shift handling with node-based Bayesian neural networks.

New algorithm robust to label corruptions in active learning.

Study finds corruption negatively impacts firm performance.

Study improves image classifier robustness to random p-norm corruptions.

Unified framework for corruption-robust linear bandits with optimal gap-dependent misspecification bounds.

We report quantitative relations between corruption level and economic factors, such as country wealth and foreign investment per capita, which are characterized by a power law spanning multiple scales of wealth and investments per capita. These relations hold for diverse countries, and also remain stable over differen…

CUTS removes corruption from models without clean data, improving utility and security.

Study robust estimation under varying corruption probabilities in data.

STAND-DA improves AD in DA target domains with limited data.

Matching the performance of conditional Generative Adversarial Networks with little supervision is an important task, especially in venturing into new domains. We design a new training algorithm, which is robust to missing or ambiguous labels. The main idea is to intentionally corrupt the labels of generated examples t…