New deep learning algorithms optimize non-decomposable measures.
problem Optimizing complex, non-decomposable performance measures.
method Directly training deep neural networks with task-specific loss functions.
result Significantly faster and more stable convergence compared to standard methods.
Meta metric learning improves few-shot learning for diverse domains.
problem Few-shot learning struggles with diverse domains and varying label numbers.
method Task-specific learners with metric learning and a meta learner to discover task-specific metrics.
result Meta metric learning achieves superior performance in diverse multi-domain tasks and flexible label numbers.
Pre-training improves model robustness and uncertainty.
problem Improving model robustness and uncertainty in machine learning.
method Adversarial pre-training and task-specific methods.
result Approximately a 10% absolute improvement in adversarial robustness.
New method improves meta-learning performance by task-specific initialization.
problem Difficulties in generalizing and achieving theoretical guarantees in conditional meta-learning.
method Structured prediction approach for task-specific initialization.
result TASML improves performance of existing meta-learning models.
Algorithm learns which weights to share in deep multi-task learning.
problem Difficulty in deciding which weights to share between tasks in deep learning models.
method Combines natural evolution strategy and stochastic gradient descent to learn optimal weight sharing.
result Task-specific networks achieve lower test errors than existing methods on multi-task learning datasets.
New algorithm preserves transport maps for better diffusion model training.
problem Training diffusion models with task-specific optimality structures.
method Generalized Schrödinger Bridge Matching (GSBM), inspired by conditional stochastic optimal control.
result GSBM better preserves transport maps, enabling stable convergence and improved scalability.
Transfer learning improves clinical time series prediction with limited data.
problem Training deep RNNs for clinical tasks requires large labeled data and tuning.
method Transfer learning from pre-trained RNNs on multiple tasks to new tasks.
result Features from pre-trained RNNs improve model performance and robustness.
Unified scoring model improves efficiency and performance across multiple tasks.
problem Efficient and resource-efficient automated scoring for diverse tasks.
method Knowledge-distilled multi-task Mixture-of-Experts (MoE) approach.
result Comparable performance to task-specific models with significantly less storage and training resources.
Proposes using gradients as features for efficient deep learning adaptation.
problem Efficient deep representation learning for different tasks.
method Designs a linear model incorporating gradients and activations of a pre-trained network.
result Shows strong results across various tasks and datasets.
UCB-TQL learns from multiple tasks with shared dynamics and adapts to task-specific variations.
problem Transfer reinforcement learning with composite MDPs where tasks share core dynamics but have sparse differences.
method UCB-TQL, a novel transfer RL algorithm for composite MDPs.
result Achieved a regret bound of i l d e O ( e H 5 N ) ilde{O}(\sqrt{eH^5N}) i l d e O ( e H 5 N ) that scales independently of the ambient dimension. AFTER technique improves NLP models by preventing overfitting to task-specific domains.
problem Standard fine-tuning degrades pretraining domain representations.
method Complements task-specific loss with adversarial objective.
result AFTER leads to improved performance on various NLP tasks.
Meta-learning approach discovers task-specific modules for few-shot tasks.
problem Meta-learning large models with limited task-specific data.
method Bayesian shrinkage for automatic discovery of task-specific and general modules.
result Outperforms existing meta-learning methods in domains with little task data and long adaptation horizons.
SWD improves unsupervised domain adaptation by measuring classifier outputs.
problem Improving unsupervised domain adaptation between domains.
method Proposes sliced Wasserstein discrepancy (SWD) for feature distribution alignment.
result SWD effectively aligns source and target distributions for various tasks.
Low-rank framework for task-specific LLM ranking from sparse comparisons.
problem Challenges in reliable task-specific ranking of LLMs under sparse, imbalanced comparisons.
method Low-rank modeling of task-by-model ability matrix, max-norm accurate estimator, task-wise top-K recovery guarantees, uncertainty quantification framework.
result Improves sample efficiency and produces tighter, better-calibrated ranking certificates.
Study on generalization in gradient-based meta-learning, showing flatter solutions and coherence between adaptation trajectories.
problem Understanding generalization in gradient-based meta-learning.
method Analysis of objective landscapes, experimental demonstration of solution properties, and empirical evidence on coherence between adaptation trajectories.
result Meta-test solutions become flatter, lower in loss, and further away from the meta-train solution as meta-training progresses, even as generalization starts to degrade.
Meta-learn Bayesian inference for task-specific BNNs using amortised inference.
problem Efficiently learning Bayesian inference for small-scale probabilistic meta-learning.
method Replace global inducing points with actual data to create a set of approximate likelihoods, train a meta-model to learn these parameters across related datasets.
result Meta-learned inference can be applied to task-specific BNNs, improving efficiency and scalability.
Paper improves natural language understanding with less data using a new training method.
problem Limited data hinders performance of small models in natural language tasks.
method Generation-Distillation: uses large finetuned models to generate new training data and distill knowledge into smaller models.
result Achieves comparable performance to BERT with 300x fewer parameters and outperforms prior distillation methods.
CompILE learns reusable segments from demonstrations for hierarchical task execution.
problem Learning reusable, variable-length segments of hierarchical behavior from demonstrations.
method Unsupervised, fully-differentiable sequence segmentation module for latent encoding and re-composition.
result Model generalizes to longer sequences and unseen environments, learns task boundaries and event encodings.
LEWIS merges LLMs without training, improving performance on specific tasks.
problem Limited performance improvement of merged models on specific benchmarks.
method Guided model merging using layer-wise sparsity and task-vector pruning.
result Improved model performance by up to 11.3% on math-solving tasks.
Graph Denoising Policy Network learns robust representations from noisy graphs.
problem Noise sensitivity in graph representation learning.
method Reinforcement learning to select signal neighborhoods and aggregate features.
result Significantly outperforms state-of-the-art methods on node classification tasks.
Enhances index selection for databases with task-specific inductive biases.
problem Challenges in traditional and automatic tuning strategies for database index set selection.
method Applies deep RL with task-specific inductive biases to index set selection, reformulating the problem as permutation learning.
result Improves index selection, achieving up to 40% smaller configurations with similar latency.
New PAC-Bayes meta-learning method improves few-shot learning accuracy and calibration.
problem Few-shot learning with limited data.
method PAC-Bayes framework extended to meta-learning, estimating task-specific posteriors.
result State-of-the-art calibration and classification results on benchmarks.
Paper proposes a pipeline for weakly supervised learning in object detection and segmentation.
problem Challenges in weakly supervised learning for object detection and segmentation with only image-level labels.
method Curriculum learning pipeline with object localization, filtering, fusing, and pixel labeling.
result State-of-the-art results in multi-label image classification and weakly supervised object detection.
CAM-GAN improves GANs for continual learning with efficient feature map transformations.
problem Efficient continual learning for GANs with reduced parameter growth.
method Designing and leveraging parameter-efficient feature map transformations, including global and task-specific parameters, residual bias, and Fisher information matrix.
result Significantly improved model performance and high-quality samples with fewer parameters.
ATAML improves text classification with attention mechanisms.
problem Limited deep learning performance with scarce data.
method Meta-learning with attention mechanisms.
result ATAML outperforms other methods on text classification tasks.
Neural networks improve clinical note representations for predictive modeling.
problem Challenges in using clinical notes for machine learning due to high dimensionality, sparsity, and scarcity of labeled data.
method Used neural networks and transfer learning to learn representations of clinical notes.
result Neural network representations significantly outperformed baseline representations in predictive models.
While sparse coding-based clustering methods have shown to be successful, their bottlenecks in both efficiency and scalability limit the practical usage. In recent years, deep learning has been proved to be a highly effective, efficient and scalable feature learning tool. In this paper, we propose to emulate the sparse…
Method selects task-specific neurons for unsupervised transfer learning.
problem Challenges in fine-tuning large language models for specific tasks.
method Selects important neurons for a specific classification task and extends to multi-source transfer learning.
result Higher similarity between task-specific fingerprints leads to better transferability.
Neural networks learn rapid adaptation with task-specific shifts.
problem Machine learning's inability to quickly adapt to new tasks.
method Conditionally shifted neurons modify activation values based on task-specific shifts.
result Models with conditionally shifted neurons achieve state-of-the-art results on metalearning benchmarks.
Meta-learning balances task-specific modeling and optimization complexity.
problem Balancing accurate task-specific modeling with ease of optimization in meta-learning.
method Theoretical and empirical analysis of trade-off between modeling and optimization in meta-learning.
result Explicit bounds on modeling and optimization errors for non-convex and linear regression problems.
New method learns disentangled representations from ECG data for VT origin classification.
problem Challenges in learning subject-specific models from ECG data due to inter-subject variations.
method Conditional Variational Autoencoder (VAE) with maximum mean discrepancy regularization and contrastive regularization.
result Demonstrated efficacy in classifying VT origin segments compared to standard VAE.
Hybrid model improves few-shot learning across diverse tasks.
problem Few-shot learning with limited data.
method Combines optimization and metric-based approaches.
result Superior performance across various settings.
LLMs improve parent-order execution in trading.
problem Improving execution costs in algorithmic trading.
method PACE (Plan-Ahead Controlled Execution) framework.
result LLMs outperform existing methods by 0.65 bps.
TDS provides exact samples for conditional distributions in diffusion models.
problem Lack of exact sampling methods for diffusion models.
method Sequential Monte Carlo (SMC) algorithm with twisting technique.
result TDS offers more accurate approximations with fewer particles compared to heuristics.
MT-net learns task-specific subspaces and metrics for gradient descent, improving few-shot learning performance.
problem Few-shot learning with gradient-based meta-learning methods.
method MT-net learns task-specific subspaces and a learned distance metric for gradient descent.
result MT-net achieves state-of-the-art performance on few-shot classification and regression tasks.
Chronos-2 forecasts multivariate and covariate data without task-specific training.
problem Limited applicability of existing time series forecasting models to real-world multivariate and covariate data.
method Chronos-2 uses a group attention mechanism for in-context learning across multiple time series.
result Chronos-2 achieves state-of-the-art performance across comprehensive benchmarks.
Plan2Explore learns new tasks efficiently through self-supervised planning.
problem Challenges in reinforcement learning, especially task-specific learning and sample efficiency.
method Self-supervised exploration and fast adaptation to new tasks through efficient planning.
result Plan2Explore outperforms prior methods in learning new tasks without supervision.
Unified MTL framework for heterogeneous data integrates shared and task-specific encoders.
problem Efficiently sharing information across multiple tasks with heterogeneous data.
method Dual-encoder framework with task-shared and task-specific encoders.
result Unified algorithm alternates learning task-specific and shared encoders and coefficients.
Reweighting training data to better represent new tasks.
problem Deploying machine learning models to new tasks is challenging due to training data distribution.
method Formulate an exponential tilt distribution shift model and learn train data importance weights to minimize KL divergence.
result The learned train data weights improve target performance evaluation, fine-tuning, and model selection.
Headless attacks bypass classification heads to fool transfer learning models.
problem Adversarial attacks against transfer learning models without access to the classification head.
method Label-blind adversarial attacks that do not require class-label information.
result Transfer attack lowers ResNet18 accuracy on CIFAR10 by over 40%.
Amobee won 3rd and 1st place in SemEval 2018 sentiment classification tasks.
problem Sentiment classification in multiple languages.
method Training GRU-CNN model with word embeddings and stacking ensembles.
result 3rd and 1st place in valence ordinal classification sub-tasks in English and Spanish.
New RL approach speeds up training across tasks.
problem Training reinforcement learning models quickly and efficiently.
method Combines planning quasi-metric and task-specific aimers.
result Achieves multiple-fold speed-up on bit-flip and robotic arm tasks.
Improves domain adaptation performance in low-resource settings.
problem Training models with scarce data in target domains.
method Augmented cyclic adversarial learning with task-specific cycle-consistency.
result Significant improvement in digit classification and speech recognition tasks.
Optimizes kernel machines using deep learning for efficiency and end-to-end learning.
problem Limited training data and lack of explicit feature sources.
method DKMO (Deep Kernel Machine Optimization) framework using Nystrom kernel approximations and deep learning.
result Improved training convergence and effectiveness over conventional model inferencing techniques.
Meta-learning improves CATE estimation from limited data.
problem Estimating heterogeneous treatment effects from scarce observational data.
method Meta-learning framework decomposes CATE estimation into sub-problems, using neural networks with shared and specific parameters, and optimizing task-specific parameters in closed form.
result Meta-learning method outperforms existing approaches in few-shot CATE estimation.
German FinBERT improves financial text analysis performance.
problem Capturing domain-specific nuances in financial text.
method German FinBERT is a pre-trained German language model trained on a large corpus of financial data.
result German FinBERT outperforms standard models on finance-specific tasks.
A tensor model for meta-learning adapts to task-specific features.
problem Learning shared representations for diverse tasks without task-specific observable information.
method Modeling meta-parameters as an order-3 tensor, estimating through tensor regression and method of moments.
result Tensor-based approach improves meta-learning performance with fewer samples.
Improved BERT model with latent persona and topic variables.
problem Improving BERT's domain-specific utility while maintaining generalization.
method Combining BERT with Universal Transformer, adding latent persona and topic variables.
result Pre-trained model for social texts outperforms baseline.