Identifies features most relevant to concept drift in data.
problem Identifying features most relevant to concept drift.
method Distinguishing between drift inducing and faithfully drifting features; deriving minimal subsets of features to characterize drift.
result Derives a detection algorithm for concept drift.
HydaLearn dynamically adjusts task weights for better MTL performance.
problem Constant loss weights in MTL lead to poor results due to drifting relevance and varying mini-batch composition.
method HydaLearn uses mini-batch gradients to dynamically adjust task weights.
result HydaLearn improves performance on synthetic and real-world data.
Language evolves over time in many ways relevant to natural language processing tasks. For example, recent occurrences of tokens 'BERT' and 'ELMO' in publications refer to neural network architectures rather than persons. This type of temporal signal is typically overlooked, but is important if one aims to deploy a mac…
This research identifies flaws in drift detection methods and creates adversarial data streams to exploit them.
problem The challenge of detecting data distribution changes (drift) in real-time systems.
method Developed adversarial data streams to show weaknesses in existing drift detection schemes.
result Demonstrated that common drift detection methods can be fooled by adversarial data streams.
Enhash detects concept drift in data streams quickly and efficiently.
problem Detecting abrupt, gradual, virtual, or recurring events in data streams.
method Uses projection hash to insert incoming samples and detects concept drift.
result Enhash has competitive performance and moderate resource requirements compared to existing ensemble learners.
This work introduces a bias-variance decomposition for proper scores, improving uncertainty estimation in predictive models.
problem Reliable uncertainty estimation for predictions in safety-critical applications, especially under domain drift.
method Developed a general bias-variance decomposition for proper scores, introducing the Bregman Information as the variance term.
result The decomposition provides novel formulations for different predictive tasks, including classification and model ensembles.
DriftMoE adapts to concept drifts in data streams efficiently.
problem Adapting to non-stationary data streams with concept drifts.
method Online Mixture-of-Experts (MoE) architecture with co-training framework.
result DriftMoE achieves competitive results with state-of-the-art stream learning adaptive ensembles.
Detects drifts in data for classification tasks using constrained embeddings.
problem Drifts in data affect model performance; unsupervised methods ignore label information.
method Task-sensitive semi-supervised drift detection with constrained low-dimensional embedding.
result Successfully detects real drifts affecting classification performance.
By appealing to renewal theory we determine the equations that the mean exit time of a continuous-time random walk with drift satisfies both when the present coincides with a jump instant or when it does not. Particular attention is paid to the corrections ensuing from the non-Markovian nature of the process. We show t…
LoRA- improves EFCL by stabilizing feature drifts.
problem Catastrophic forgetting in exemplar-free continual learning.
method LoRA-subtraction to adaptively subtract old task weights.
result LoRA- achieves state-of-the-art performance in EFCL.
A new SNN model explains decision-making with learning and spiking neurons.
problem Lack of learning mechanism in existing models for decision-making.
method Proposes a Spiking Neural Network (SNN) model that incorporates a learning mechanism and uses multivariate Hawkes processes.
result Shows a coupling between DDM and Poisson counter models and derives a DDM from a Hawkes network of spiking neurons.
New framework tackles DG under posterior drift, where optimal classifier varies by domain.
problem Generalizing from multiple domains with varying optimal classifiers.
method Decision-theoretic framework for DG under posterior drift.
result Optimal classifier can vary significantly across domains, challenging existing DG approaches.
A new approach switches between simple and complex models to handle concept drifts in regression tasks.
problem Handling concept drifts in regression models to maintain accurate predictions over time.
method Error Intersection Approach: switches between simple and complex models based on drift detection.
result The Error Intersection Approach significantly outperforms baselines in handling concept drifts in a real-world taxi demand dataset.
Adaptive sampling detects local concept drift with limited labels.
problem Detecting local concept drift in dynamic environments with scarce labels.
method Combines residual-based exploration and exploitation with EWMA monitoring.
result Superior performance in label efficiency and drift detection accuracy.
This research generates synthetic data streams for handling concept drifts and novel classes.
problem Handling concept drifts and novel classes in dynamic data streams.
method Synthetic data stream generation for both concept drifts and novel classes.
result Demonstrates the effectiveness of unsupervised drift detectors in open set recognition.
This review covers learning under concept drift, including detection, understanding, and adaptation.
problem Unforeseeable changes in data distribution over time impact machine learning performance.
method Reviews and analyzes methodologies and techniques for concept drift detection, understanding, and adaptation.
result Establishes a framework for learning under concept drift with three main components.
New budget quantifies drift in closed-loop learning, improving reproducibility.
problem Characterizing statistical learning under distributional drift in closed-loop settings.
method Introduces an intrinsic drift budget CT quantifying cumulative information-geometric motion of the data distribution. result Proves a drift-feedback bound of order T−1/2+CT/T for prequential reproducibility, up to controlled second-order remainder terms. Data-driven predictive analytics are in use today across a number of industrial applications, but further integration is hindered by the requirement of similarity among model training and test data distributions. This paper addresses the need of learning from possibly nonstationary data streams, or under concept drift,…
NP-PROV separates mean and variance spaces to improve function uncertainty.
problem Neural Processes fail on out-of-domain tasks due to shared latent space uncertainty.
method Separates mean and variance into function-value-related and position-related latent spaces.
result NP-PROV achieves state-of-the-art likelihood with bounded variance in drifts.
New method detects when models influence their own drift in real-time data streams.
problem Models can induce concept drift in real-time data streams.
method CheckerBoard Performative Drift Detection (CB-PDD)
result CB-PDD effectively detects performative drift in real-time data streams.
Geometric stability predicts steerability and detects drift in language models.
problem Predicting steerability and detecting drift in language models.
method Supervised and unsupervised geometric stability measures.
result Supervised geometric stability predicts steerability with high accuracy and detects drift earlier.
This paper uses SLT to ensure learning guarantees in CD detection.
problem Lack of learning guarantees in CD detection algorithms.
method Adapting SLT assumptions to CD scenarios to ensure learning guarantees.
result Ensured learning guarantees in CD detection algorithms.
Classifiers operating in a dynamic, real world environment, are vulnerable to adversarial activity, which causes the data distribution to change over time. These changes are traditionally referred to as concept drift, and several approaches have been developed in literature to deal with the problem of drift handling an…
A new method detects concept drift without true labels.
problem Detecting concept drift in unsupervised settings.
method Student-teacher learning paradigm for drift detection.
result The method outperforms state-of-the-art approaches in experiments.
A drift detection method for large datasets without labels.
problem Early detection of concept drift in large, unlabeled datasets.
method Classical statistical process control in a label-less setting.
result Better statistical power than previous methods under computational constraints.
Develops PromptShift-CRC for drift-aware conformal risk control in foundation models under prompt and domain shift.
problem Fixed calibration risk in foundation models due to prompt and domain shift.
method Embeds prompts and responses, measures drift, gives more weight to recent examples, and updates risk online.
result Develops method to control risk up to terms for distribution mismatch and weighted quantile uncertainty.
Study explains mortgage burnout using Cox hazard models.
problem Understanding burnout in mortgage pools.
method Modeling mortgage prepayment using Cox hazard processes.
result Observed pool hazard is a survival-weighted mean of individual hazards with a selection term.
DRIFT uses RL to automate functional software testing efficiently.
problem Efficient and reliable automated software testing.
method DRIFT employs Q-learning with Graph Neural Networks on symbolic UI representations.
result DRIFT can robustly test software functionalities in a fully automated manner.
LLMs cause inconsistent financial outputs, smaller models are more reliable.
problem Inconsistent outputs from LLMs undermine auditability and trust in financial workflows.
method Finance-calibrated deterministic test harness, task-specific invariant checking, model classification, and cross-provider validation.
result Smaller models (Granite-3-8B, Qwen2.5-7B) achieve 100% output consistency, while larger models like GPT-OSS-120B have high drift.
This paper tackles continuous domain adaptation with a new approach.
problem Learning in non-stationary environments, especially domain drift.
method Variational domain-agnostic feature replay, composed of inference, generative, and solver modules.
result Demonstrates the effectiveness of the proposed approach for practical usage.
We study the Markowitz portfolio selection problem with unknown drift vector in the multidimensional framework. The prior belief on the uncertain expected rate of return is modeled by an arbitrary probability law, and a Bayesian approach from filtering theory is used to learn the posterior distribution about the drift …
A new method detects concept drift in streaming data using k-means space partitioning.
problem Detecting distribution changes in streaming data.
method Equal intensity k-means space partitioning (EI-kMeans) and heuristic sensitivity improvement.
result EI-kMeans improves drift detection accuracy and sensitivity.
Framework models supervised learning in non-stationary data.
problem Non-stationary data in supervised learning.
method Statistical physics methods applied to LVQ and neural networks.
result LVQ and ReLU have different sensitivity to concept drift.
LUNAR uses cellular automata for real-time data classification in fast streams.
problem Real-time machine learning challenges with fast data streams and concept drift.
method Streamified cellular automata approach for incremental learning and adaptation.
result Competitive performance in classification compared to established online learning methods.
Classifiers deployed in the real world operate in a dynamic environment, where the data distribution can change over time. These changes, referred to as concept drift, can cause the predictive performance of the classifier to drop over time, thereby making it obsolete. To be of any real use, these classifiers need to d…
ZDP detects drift in large language models without labels, proving key theorems and metrics.
problem Detecting drift in large language models without task labels or output evaluations.
method Zero-Direction Probing (ZDP) framework based on null directions of transformer activations, proving theoretical guarantees.
result Proves the Variance--Leak Theorem, Fisher Null-Conservation, Rank--Leak bound, and logarithmic-regret guarantee.
stream-learn is a Python library for analyzing data streams with various drift types.
problem Analyzing drifting and imbalanced data streams.
method Synthetic data stream generator, evaluation methodologies, and imbalanced binary classification metrics.
result Efficient implementation of classifiers for data stream analysis.
DDG-DA predicts future data distribution to adapt models for predictable concept drift.
problem Adapting models to streaming data with predictable concept drift.
method Train a predictor to forecast future data distribution, generate training samples, and train models on them.
result Significant improvement on multiple models in real-world tasks.
New method identifies drift and diffusivity from SDE marginals.
problem Challenging task to identify drift and diffusion from SDE population dynamics.
method Proposes nn-APPEX, a Schrodinger Bridge-based inference method.
result Gradient-flow drift and Brownian diffusivity jointly identifiable from marginals.
New method identifies SDE drift and diffusion from temporal data.
problem Learning SDE parameters from temporal data, especially in noisy or incomplete data.
method Entropy-regularized optimal transport, APPEX algorithm.
result Can almost always recover drift and diffusion from temporal marginals.
Paper presents IMRCs for evolving tasks with forward and backward learning.
problem Incremental learning of evolving tasks with few samples per task.
method Incremental minimax risk classifiers (IMRCs) that exploit forward and backward learning.
result IMRCs provide significant performance improvement, especially with reduced sample sizes.
Wealth redistribution through Fokker-Planck equation controls preserves Gini coefficient.
problem Preserving Gini coefficient through proportional wealth tax.
method Formulating optimal redistribution as a control problem for Fokker-Planck equation.
result Progressive taxes redistribute within policy-relevant timescales.
Motivated by empirical data, we develop a statistical description of the queue dynamics for large tick assets based on a two-dimensional Fokker-Planck (diffusion) equation, that explicitly includes state dependence, i.e. the fact that the drift and diffusion depends on the volume present on both sides of the spread. "J…
ToolChain-CRC addresses the risk-control problem for retrieval-augmented and tool-using agents under drift.
problem Risk-control problem for retrieval-augmented and tool-using agents under drift.
method ToolChain-CRC uses conformal risk-control under exchangeable calibration runs.
result Trajectory-level risk control keeps accepted-trajectory risk below the target.
A geometric theory explains loss functions for robust representation learning.
problem Treats robustness, domain adaptation, and sensor drift as separate literatures.
method Estimates covariance Sigma_task and uses it to pin Jacobian penalties.
result Proves optimality and necessity of range coverage for penalty matrices.
ProteuS generates synthetic financial data with regime changes for testing drift detection.
problem Simulating concept drift in financial markets for model evaluation.
method ARMA-GARCH models fitted to ETF data, generating synthetic time series with predefined regime changes.
result Generated datasets reveal the complexity of detecting and adapting to market regime changes.
TALBO optimizes latent spaces for evolving design objectives.
problem Temporal drift in design objectives.
method GP-prior variational autoencoder for time-varying latent space.
result Consistently outperforms LSBO baselines across varying drift speeds and objectives.
This paper extends neural network approximation results to denoising diffusion models.
problem Improving the efficiency and accuracy of generative models.
method Leveraging connections to stochastic control and neural network approximation.
result Established neural network approximation results for the Föllmer drift are extended to denoising diffusion models.