Deep ensembles don't necessarily improve calibration in low data regimes.
problem Calibration issues in deep learning models, especially in low data regimes.
method Examination of data-augmentation, ensembling, and post-processing calibration methods.
result Standard ensembling techniques can lead to less calibrated models in low data regimes.
Adding uninformative labels improves tumor segmentation in low-data mammography.
problem Improving tumor segmentation in mammography with limited data.
method Used seemingly uninformative labels from non-expert annotators to turn a multi-label task into a multi-class problem.
result Performance gains in tumor segmentation are achieved in low-data settings with additional uninformative labels.
Enhances classification accuracy on low data sets using synthetic data.
problem Low sample size in data augmentation.
method Variational Autoencoder and manifold sampling.
result Significant improvement in classification accuracy (e.g., 88.6% vs 80.7%).
Efficiently learns 3D convolutions with less data.
problem High parameter and data costs in 3D convolutions.
method Temporal factorization of 3D kernels.
result Significantly reduces training data requirement and parameter count.
This study compares MLPs and KANs in low-data regimes, finding MLPs with personalized activation functions outperform KANs.
problem Comparing MLPs and KANs in low-data regimes.
method Introduced an effective technique for designing MLPs with unique, parameterized activation functions for each neuron.
result MLPs with personalized activation functions achieve significantly higher predictive accuracy with only a modest increase in parameters, especially in low-data regimes.
A new method for anomaly detection adapts to local non-stationarity in low-data regimes.
problem Adapting conformal anomaly detection to handle distribution shifts in real-world data.
method Proposes a continuous inference relaxation using continuous weighted kernel density estimation to decouple local adaptation from tail resolution.
result Restores detection capabilities and statistical power in low-data regimes while maintaining valid error control.
Improved computed tomography reconstruction with deep learning and deep image prior.
problem Low data efficiency in computed tomography reconstruction.
method Combining learned primal-dual methods with deep image prior for improved quality and generalization.
result Proposed methods outperform state-of-the-art in low data regime.
Proposes GPLFR for predicting high-dimensional outputs with few data.
problem Predicting high-dimensional outputs from limited data.
method GPLFR combines Gaussian process and linear-Gaussian decoding for high-dimensional prediction.
result GPLFR outperforms existing methods in predicting high-dimensional outputs.
Stein discrepancy improves UDA performance in low-data scenarios.
problem Improving model performance on unlabeled target domains with limited data.
method Proposes a novel UDA framework using Stein discrepancy, an asymmetric measure that depends on the target distribution through its score function.
result Consistently outperforms prior UDA approaches under limited target data across multiple benchmarks.
Ensemble pre-trained models for low data transfer learning.
problem Training good models from scratch in low data regime.
method Fine-tune nearest-neighbour ranked pre-trained models to create ensembles.
result Achieves state-of-the-art performance with lower inference budget.
The paper proposes a machine learning framework for portfolio optimization with limited data.
problem Low data environments and regime uncertainty in portfolio optimization.
method A teacher-student learning pipeline with CVaR optimizer generating supervisory labels and neural models trained on real and synthetic data.
result Student models can match or outperform the CVaR teacher and achieve improved robustness under regime shifts.
We train a network to generate mappings between training sets and classification policies (a 'classifier generator') by conditioning on the entire training set via an attentional mechanism. The network is directly optimized for test set performance on an training set of related tasks, which is then transferred to unsee…
New framework improves fairness in small data settings.
problem Ensuring fairness in low-data environments.
method Combines posterior sampling exploration with fair classification.
result Framework maximizes accuracy while meeting fairness constraints.
This study improves UAV identification using RF signals with one-shot generative methods.
problem Limited RF environments and signal variability make traditional RF identification ineffective.
method Introduces one-shot generative methods to augment RF signals for UAV identification.
result One-shot generative methods outperform traditional methods in low-data regimes.
Kernelised flows improve density estimation and generation with fewer parameters.
problem Limited expressiveness of flow-based models due to invertibility constraints.
method Integrates kernels into normalising flows to enhance expressiveness and efficiency.
result Kernelised flows outperform neural network-based flows in parameter efficiency and low-data scenarios.
BCD Nets use variational inference to estimate DAGs with uncertainty.
problem Uncertainty in inferring causal graphs from limited data.
method Variational inference framework for Bayesian DAG estimation.
result BCD Nets outperform maximum-likelihood methods in low data regimes.
Study compares deep feature methods for anomaly detection in limited data scenarios.
problem Handling limited data in industrial inspection applications.
method Three approaches (KNN, Mahalanobis, PaDiM) using pre-trained deep features with data augmentation.
result Data augmentation significantly improves performance in small data regimes.
Effective training of neural networks requires much data. In the low-data regime, parameters are underdetermined, and learnt networks generalise poorly. Data Augmentation alleviates this by using existing data more effectively. However standard data augmentation produces only limited plausible alternative data. Given t…
Bayesian autoencoders discover physics from noisy data.
problem Challenges in identifying governing equations and coordinates from noisy, low-data real-world data.
method Bayesian SINDy autoencoders with hierarchical Bayesian sparsifying prior and adaptive empirical Bayesian method.
result Better physics discovery with lower data and fewer training epochs, along with valid uncertainty quantification.
TempBalance boosts model performance with low data.
problem Low-data training and fine-tuning in model alignment.
method Inspired by HT-SR theory, TempBalance balances training quality across layers.
result TempBalance improves model performance as data decreases.
pFedGP uses Gaussian processes for personalized federated learning with improved kernel learning.
problem Learning personalized models across clients with limited data.
method pFedGP uses Gaussian processes with deep kernel learning, including a shared neural network kernel and inducing points.
result pFedGP achieves well-calibrated predictions and significantly outperforms baseline methods.
New learning rules for wide neural networks without backpropagation.
problem Training wide neural networks efficiently and without backpropagation.
method Input-weight alignment driven by gradient descent in the NTK regime.
result Biologically-motivated learning rules equivalent to backpropagation in wide networks.
Today, the dominant paradigm for training neural networks involves minimizing task loss on a large dataset. Using world knowledge to inform a model, and yet retain the ability to perform end-to-end training remains an open question. In this paper, we present a novel framework for introducing declarative knowledge to ne…
Method estimates multiple related Gaussian distributions using Laplacian regularization.
problem Jointly estimate multiple related zero-mean Gaussian distributions.
method Laplacian regularized stratified model fitting with hyper-parameters to encourage covariance closeness.
result The method performs well, especially in low data regimes, as demonstrated in finance, radar, and weather.
We propose a new class of probabilistic neural-symbolic models, that have symbolic functional programs as a latent, stochastic variable. Instantiated in the context of visual question answering, our probabilistic formulation offers two key conceptual advantages over prior neural-symbolic models for VQA. Firstly, the pr…
Gradient-based meta-learning techniques are both widely applicable and proficient at solving challenging few-shot learning and fast adaptation problems. However, they have practical difficulties when operating on high-dimensional parameter spaces in extreme low-data regimes. We show that it is possible to bypass these …
The supervised learning paradigm is limited by the cost - and sometimes the impracticality - of data collection and labeling in multiple domains. Self-supervised learning, a paradigm which exploits the structure of unlabeled data to create learning problems that can be solved with standard supervised approaches, has sh…
Geometric interpretation improves VAE performance and robustness.
problem Improving Variational Autoencoder performance and robustness.
method Introducing a geometric perspective on VAEs, sampling from the Riemannian latent space.
result Improved generation and interpolations with competitive or better performance on benchmark datasets.
Deep reinforcement learning requires a heavy price in terms of sample efficiency and overparameterization in the neural networks used for function approximation. In this work, we use tensor factorization in order to learn more compact representation for reinforcement learning policies. We show empirically that in the l…
Manipulating data, such as weighting data examples or augmenting with new instances, has been increasingly used to improve model training. Previous work has studied various rule- or learning-based approaches designed for specific types of data manipulation. In this work, we propose a new method that supports learning d…
Improves feature selection in high-dimensional data using LLM-generated weights.
problem Inaccurate LLM-generated weights degrade feature selection performance.
method Integrates LLM-generated weights into prior inclusion probabilities using LLM Sparsity Prior (LSP).
result Improves prediction accuracy and identifies clinically relevant features.
Study improves posterior inference in neural processes with limited data.
problem Improving posterior predictive inference in probabilistic models with scarce conditioning data.
method Examined effects of pooling operators and variational families on posterior quality in neural processes.
result Novel neural process architectures lead to superior posterior predictive samples in image completion/in-painting tasks.
New algorithm improves matrix estimation with one-sided covariates.
problem Estimating matrix means with unobserved row covariates.
method Proposes an algorithm for nonparametric matrix estimation with observed column covariates.
result Achieves minimax optimal nonparametric rate in moderately proportioned matrices.
DOODL learns shared spectral dynamics across related dynamical systems.
problem Learning independent dynamical operators for each system limits discovery of shared structure.
method DOODL learns a dictionary of characteristic spectral dynamics on a manifold of related systems.
result DOODL achieves errors one to two orders of magnitude lower than independent operator estimation methods.
GIT uses gradient estimators to target interventions for causal discovery.
problem Challenges in inferring causal structure from observational data.
method GIT uses gradient estimators to target interventions for causal discovery.
result GIT performs on par with competitive baselines, surpassing them in low-data regimes.
PILLAR improves SP learning with less private data.
problem Efficiently learning with semi-private data under privacy constraints.
method Uses pre-trained public data features to reduce private data requirements.
result Significantly lower private labelled sample complexity achieved.
EHVI outperforms scalarized EI in MOBO for molecule design.
problem Benchmarking MOBO strategies for molecule design.
method Compared EHVI against fixed-weight scalarized EI in MOBO.
result EHVI consistently outperforms scalarized EI in molecular optimization tasks.
Paper tackles robustness in adversarial noise with a meta-optimizer.
problem Sensitivity to adversarial noise hinders machine learning deployment.
method Meta-optimizer learns to robustly optimize models using adversarial examples.
result Meta-optimizer transfers adversarial knowledge to new models without generating new examples.
Develops a method to ensure accuracy of few-shot transfer learning models.
problem Lack of generalization guarantees for low-data transfer learning.
method Trains a distribution over PEFT parameters using upstream tasks and samples plausible PEFTs for downstream tasks.
result Demonstrates non-vacuous generalization guarantees compared to existing methods in the low-shot regime.
Bayesian VAR model discovers Granger causality with uncertainty-aware binary graphs.
problem Discovering Granger causal relations from multivariate time-series data.
method Bayesian Vector AutoRegression with factorised Granger-Causal Graphs.
result Our method achieves better performance, especially in low-data regimes.
AuxiLearn combines auxiliary tasks into a single loss function.
problem Improving neural network performance on a main task using auxiliary tasks.
method Implicit differentiation to learn a network that combines auxiliary tasks into a single coherent objective function.
result AuxiLearn consistently outperforms competing methods in various tasks and domains.
Traditional text classifiers are limited to predicting over a fixed set of labels. However, in many real-world applications the label set is frequently changing. For example, in intent classification, new intents may be added over time while others are removed. We propose to address the problem of dynamic text classifi…
XIMP improves molecular property prediction by integrating multiple graph representations.
problem Graph neural networks struggle in data-scarce regimes and fail to surpass traditional methods.
method Cross-graph inter-message passing with multiple graph abstractions.
result XIMP outperforms state-of-the-art baselines across diverse molecular property tasks.
We study the challenges of applying deep learning to gene expression data. We find experimentally that there exists non-linear signal in the data, however is it not discovered automatically given the noise and low numbers of samples used in most research. We discuss how gene interaction graphs (same pathway, protein-pr…
Task-agnostic data augmentation shows little benefit for pretrained transformers.
problem Evaluating the effectiveness of task-agnostic data augmentation on pretrained transformers.
method Conducted a systematic examination of two data augmentation techniques (Easy Data Augmentation and Back-Translation) across 5 tasks, 6 datasets, and 3 pretrained transformer models.
result Data augmentation techniques previously effective for non-pretrained models fail to consistently improve performance for pretrained transformers, even with limited training data.
Adaptive method improves prediction intervals with global coverage guarantees and local error distribution.
problem Global coverage guarantees of conformal regression are often violated by local error distributions.
method Adaptive Conformal Regression with Jackknife+ Rescaled Scores
result Improves local coverage without sacrificing global coverage, especially in low-data regimes.
MFNets constructs efficient multifidelity surrogates from diverse information sources.
problem Creating accurate surrogates from multiple, potentially costly or inaccurate data sources.
method Directed acyclic graph of connections, gradient-based minimization of least squares objective, flexible information source structure.
result Error reduction by orders-of-magnitude, especially in low-data scenarios.
ScaML-GP efficiently learns from few meta-tasks using Gaussian processes.
problem Exploiting historical data for quick task solving in low-data regimes.
method Modular Gaussian process model with a carefully designed multi-task kernel.
result ScaML-GP learns efficiently with few and many meta-tasks.