New RL approach transfers policies across related domains.
problem Efficiently transfer policies between different domains in RL.
method Adapts and reuses optimal policies from related source tasks.
result Improves sample efficiency in target domain learning.
Paper proposes a method to identify negative transfers in multitask learning using surrogate models.
problem Identifying subsets of source tasks that improve target task performance in multitask learning.
method Surrogate modeling to precompute multitask learning performances and approximate them with a linear regression model.
result The approach predicts negative transfers from multiple source tasks to target tasks more accurately than existing methods.
Paper presents MTTDSC for better target-specific sentiment classification.
problem Improving accuracy in detecting and aggregating sentiments towards specific targets in social media.
method MTTDSC uses a multi-task learning approach with an auxiliary task for passage-level sentiment classification and a main task for target-specific sentiment classification.
result MTTDSC outperforms state-of-the-art baselines in sentiment classification.
Trains recurrent networks to perform complex tasks with fewer neurons and better robustness.
problem Training recurrent networks to handle temporally complex tasks efficiently.
method Introduces a target network to provide suitable dynamics for the task, modifying the full connectivity matrix.
result Trained networks perform tasks with fewer neurons and greater noise robustness.
Proposes a method to transfer samples from source tasks to target tasks in RL.
problem Improving RL learning by selecting and weighting relevant samples from multiple tasks.
method Automatic estimation of importance weights for each source sample, applied to a batch RL algorithm.
result The proposed method achieves better learning performance and robustness to task differences.
Novel hyperparameter optimization for target tasks under covariate shift.
problem Hyperparameter optimization under multi-source covariate shift.
method Construct variance reduced estimator to unbiasedly approximate target objective; propose no-regret hyperparameter optimization procedure.
result Proposed framework broadens applications of automated hyperparameter optimization.
Method transfers knowledge without label overlap, source data, or target architecture consistency.
problem Difficulties in transfer learning due to label mismatch, restricted source data, and specialized target architectures.
method Uses deep generative models in two stages: pseudo pre-training and pseudo semi-supervised learning.
result Outperforms scratch training and knowledge distillation methods.
TASFAR adapts regression models without labeled source data.
problem Lack of labeled source data for domain adaptation.
method Uses prediction confidence to estimate target label distribution and calibrate source model.
result Substantially reduces errors in various regression tasks.
TarMAC targets and coordinates multi-agent communication for cooperative tasks.
problem Coordinating multi-agent reinforcement learning in partially observable environments.
method Targeted multi-round communication approach without supervision.
result Improved performance and sample efficiency in diverse environments.
Double descent in transfer learning explained for linear regression problems.
problem Understanding generalization errors in transferring parameters between overparameterized linear regression tasks.
method Analytical characterization of generalization error in terms of transfer learning factors.
result Generalization error follows a two-dimensional double descent trend controlled by transfer learning factors.
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.
A new method for task transfer in reinforcement learning using expert preferences.
problem Inconveniently obtaining expert demonstrations and cost functions for task transfer.
method Develops a novel framework that uses expert preferences to select relevant demonstrations and learns the target cost function and trajectory distribution.
result Demonstrated effectiveness through simulations on various benchmarks.
Proposes a new Q-learning method to improve sample complexity.
problem Improving sample complexity in Q-learning with limited data.
method Integrates Q-function from a source task into a target task under safe conditions.
result The method converges faster than standard Q-learning under certain conditions.
RL policy tracks dynamic targets in partially known environments robustly.
problem Active target tracking in partially known environments.
method Deep reinforcement learning (RL) approach for in-sight tracking, navigation, and exploration.
result Unified RL policy shows robust behavior for agile and anomalous targets.
A new meta-learning framework that assigns weights to source tasks based on target samples.
problem Learning initialization for target tasks with limited labeled examples.
method A general framework that assigns weights to the loss of different source tasks, which can depend on the target samples. Provides upper bounds and develops a learning algorithm based on minimizing the error bound with respect to an empirical IPM.
result Empirically, the weighted meta-learning algorithm finds better initializations than uniformly-weighted meta-learning algorithms.
Few-shot learning improves time-series forecasting with limited data.
problem Limited data in target tasks degrade forecasting performance.
method A few-shot learning method using recurrent neural networks with attention.
result The model forecasts future values effectively with minimal data.
New distance metric for neural architecture search reduces search space complexity.
problem Reducing the complexity of neural architecture search.
method Fisher task distance for measuring task similarity and online neural architecture search.
result Reduced search space complexity for task-specific architectures.
RTE enables extrapolation to new tasks by learning task transformations.
problem Learning systems struggle to generalize to unseen tasks.
method Relational Task Extrapolator (RTE) learns task transformations to enable extrapolation.
result RTE substantially outperforms existing approaches on extrapolation tasks.
Study on neural networks' performance in sequential task learning.
problem Understanding the performance of neural networks in sequential task learning.
method Theoretical analysis of generalization performance in continual learning using statistical mechanical analysis of kernel ridge-less regression.
result Characteristic transitions from positive to negative transfer observed in neural networks.
The study analyzes transfer learning in infinite-width neural networks, improving generalization on target tasks.
problem Improving generalization in neural networks when using pretraining on a source task.
method Developed a theory under gradient flow for infinitely wide networks, analyzing fine-tuning and joint pretraining.
result Summary statistics of randomly initialized networks after pretraining are adaptive kernels that depend on both source and target data.
Proposes a novel MTL approach based on bias-variance analysis.
problem Improving multi-task learning performance through shared knowledge.
method Two-phase iterative aggregation of targets and features using bias-variance analysis.
result Validation on synthetic and real-world datasets demonstrates the effectiveness of the proposed method.
New findings on representation learning beyond linear functions.
problem Achieving diversity in representation learning beyond linear prediction functions.
method Analysis of eluder dimension and empirical risks.
result Diversity holds even with nonlinear prediction functions and multiple layers in neural networks.
A model learns set comparison, vague quantification, and proportional estimation from visual scenes.
problem Learning of quantification mechanisms from visual scenes.
method A multi-task computational model that integrates set comparison, vague quantification, and proportional estimation.
result The multi-task model performs better when lower-complexity tasks are available and can generalize to unseen combinations.
Improves score estimation for noised targets using known clean scores.
problem Poor score estimation at low noise levels in Denoising Score Matching.
method Introduces Target Score Identity and Target Score Matching loss.
result Score estimates are more accurate at low noise levels.
One-step Bellman alignment improves online RL by reducing task mismatch.
problem Online RL struggles with task similarity defined by rewards or transitions.
method One-step Bellman alignment and re-weighted targeting (RWT) to correct task mismatch.
result Regret bounds show task shift complexity, not target MDP, affects performance.
Domain Fusion uses GANs to augment data for low-volume target datasets.
problem High costs in data development for deep learning applications.
method Multi-domain learning GANs to generate new samples.
result Domain Fusion achieves better classification accuracy with less data.
SCA learns to learn from target-set examples, improving few-shot learning performance.
problem Limited labelled examples in few-shot learning tasks.
method SCA learns a label-free loss function via self-critique and adapts the base-model for new tasks.
result SCA achieves substantially reduced error-rates and state-of-the-art performance.
Proposes a method to directly classify target domain samples without labeled data.
problem Directly classifying unlabeled target domain samples in domain adaptation.
method Trains a classifier to classify both source and target domain samples unsupervisedly.
result Achieves state-of-the-art results in unsupervised domain adaptation tasks.
Proposes a new self-taught learning approach for SVM classification.
problem Classification with limited labeled target data and abundant unlabeled source data.
method Introduces a new objective function that learns a common space and robust SVM classifiers.
result Superior performance compared to existing algorithms on Caltech-256 and MSRC+LMO datasets.
DEN learns diverse tasks to generalize to unseen tasks.
problem Generalization from a diverse set of classification tasks with limited data.
method Three-block architecture: covariate transformation, distribution embedding, and classification.
result DEN outperforms existing methods in various synthetic and real tasks.
New theory explains when pre-trained models can improve downstream tasks.
problem Lack of theoretical understanding of task similarity for transfer learning.
method Feature-centric viewpoint, theoretical results on transferability phase diagram.
result Transfer learning outperforms training from scratch when target task is well represented in feature space.
Improves neural network regression performance using encoder-decoder approach.
problem Suboptimal regression results from classical methods in neural networks.
method Proposes a new method using an encoder-decoder approach for regression.
result Shows improved performance on various real-world datasets.
Improves AI agents' 3D navigation by learning from failures and 3D spatial relationships.
problem Challenges in data efficiency, obstacle avoidance, and generalization in 3D visual navigation.
method Incorporates attention on 3D spatial relationships and a target skill extension module into DRL framework.
result Significantly improves navigation performance and generalization across targets and scenes.
Improved model training for few-class, few-shot tasks.
problem Meta-learning algorithms struggle in many-shot and many-class settings.
method Joint training approach combining transfer-learning and meta-learning.
result Improved generalization performance on unseen tasks.
SpotTune adapts fine-tuning strategies per instance for improved transfer learning.
problem Improving transfer learning performance with deep neural networks.
method Adaptive fine-tuning approach using policy networks to decide whether to use pre-trained or fine-tuned layers.
result SpotTune outperforms traditional fine-tuning on 12 out of 14 standard datasets and achieves highest scores on Visual Decathlon.
XMixup improves transfer learning accuracy by 1.9% with less training time.
problem Efficiently transfer knowledge from large source datasets to target tasks with small samples.
method Cross-domain Mixup technique that selects auxiliary samples from source datasets and augments training samples via mixup strategy.
result Improves accuracy by 1.9% on average over six real-world transfer learning datasets.
The paper tackles HS target localization using robust PCA with dictionary-based approach.
problem Localized target detection in hyperspectral images.
method Formulates HS image as low-rank + dictionary sparse, develops recovery guarantees.
result Recovery guarantees and performance analysis on real HS datasets.
Auxiliary Tuning adapts pre-trained models for novel tasks efficiently.
problem Adapting pre-trained models for new tasks efficiently.
method Supplementing pre-trained model with an auxiliary model that shifts output distribution.
result Achieved similar results to training from scratch with fewer resources.
MRTL transfers knowledge across multiple target domains using shared latent factors.
problem Transfer learning in label-scarce target domains.
method MRTL uses collective nonnegative matrix tri-factorization to transfer knowledge from multiple sources to multiple targets.
result MRTL achieves better performance than state-of-the-art methods.
Paper tackles domain adaptation without labeled target data.
problem Performing well on an unlabeled target domain using only labeled source data.
method Learning self-supervised tasks on both source and target domains simultaneously.
result Successfully generalizes to the unlabeled target domain.
This paper proposes Dropping Networks for improved transfer learning in natural language understanding tasks.
problem Transfer learning between natural language understanding tasks often suffers from negative transfer.
method Combines Dropout and Bagging (Dropping) for improved transferability in neural networks.
result Improves transfer learning performance and comparable results to state-of-the-art using a fraction of target task data.
This work transfers causal knowledge between tasks for Individual Treatment Effect estimation.
problem Estimating Individual Treatment Effects (ITE) requires a large amount of data, making it challenging.
method The authors introduce a practical framework for efficient transfer of causal knowledge between tasks, using a Causal Inference Task Affinity (CITA) measure.
result ITE knowledge transfer can significantly reduce the amount of data needed for ITE estimation.
HIDRA optimizes weights for diverse tasks, improving model performance.
problem Optimizing gradient-based optimization strategies with dynamic target variables.
method Meta-learning approach that learns a master neuron to initialize output neurons for any number of target variables.
result Improves model performance and generalizes to tasks with any number of target variables.
NAM maps images between domains without adversarial learning, achieving better quality and stability.
problem Translating images between domains without correspondences and adversarial learning.
method NAM separates generative modeling and cross-domain mapping, using a pre-trained target model.
result Higher quality and resolution image translations with simpler and more stable training.
CUDA CTDR tackles unsupervised domain adaptation without domain alignment.
problem Lack of direct methods for unlabeled target domain classification.
method Jointly learns CTDR on source and target distributions using contradistinguish loss and supervised loss.
result CUDA CTDR achieves state-of-the-art results on various domain adaptation datasets.
Paper defines a new distance metric for comparing learning tasks.
problem Comparing difficulty of learning tasks between source and target.
method Information geometry, optimal transport, coupled transfer distance.
result Coupled transfer distance correlates with fine-tuning difficulty.
TL-PCA uses transfer learning to improve PCA performance with limited target data.
problem PCA performance is limited with scarce target data.
method Transfer learning approach to PCA (TL-PCA) that combines source task knowledge with target task data.
result Improved PCA representation for dimensionality reduction with limited target data.
Efficiently optimize expensive functions using auxiliary task information.
problem Optimizing expensive functions with limited data.
method Gaussian process with neural network mean and covariance functions, leveraging auxiliary task information.
result Identifies optimal points with fewer evaluations than existing methods.