A new method detects changes in machine learning models over time.
problem Automatic monitoring of machine learning models trained on evolving data.
method Score-based statistical hypothesis test for change detection.
result The method can detect changes in any number of model components.
Study uses property elicitation to understand how fairness regularizers affect optimal decisions.
problem Understanding how fairness regularizers change the optimal decision in predictive algorithms.
method Property elicitation to analyze the relationship between loss, regularization, and optimal decision.
result Necessary and sufficient condition for when a property changes with the addition of a regularizer.
DeltaGrad rapidly retrain models with minimal data changes.
problem Rapid retraining of machine learning models with minimal data changes.
method DeltaGrad algorithm based on cached training information.
result DeltaGrad compares favorably to state-of-the-art methods.
Reduces change detection to estimation using confidence sequences.
problem Detecting changes in data streams with minimal delay and false alarms.
method Reduction from sequential change detection to sequential estimation using confidence sequences.
result Change detection scheme with minimal structural assumptions and strong guarantees.
Method identifies change points in high-dimensional models using sample weights.
problem Identifying change points in high-dimensional generalized linear models.
method Sample-weighted empirical risk minimization (Weighted ERM).
result Weighted ERM yields precise asymptotic performance characterization for Gaussian designs.
AdaRL adapts quickly to new environments with minimal data.
problem Quickly adapting to new environments in reinforcement learning.
method AdaRL uses a parsimonious graphical representation to encode changes across domains.
result AdaRL can efficiently adapt policies to target domains with few samples.
The change in Holographic entanglement entropy (HEE) for small fluctuations about pure anti De Sitter (AdS) is obtained by a perturbative expansion of the area functional in terms of the change in the bulk metric and the embedded extremal surface. However, it is known that change in the embedding appears in second orde…
Positive braids minimize knot untangling steps.
problem Finding the minimum number of steps to untangle knots.
method Analyzing positive braids and their knot closures, comparing ascending number to unknotting number.
result Ascending number equals unknotting number for knots from positive braids.
A framework previously introduced in [3] for solving a sequence of stochastic optimization problems with bounded changes in the minimizers is extended and applied to machine learning problems such as regression and classification. The stochastic optimization problems arising in these machine learning problems is solved…
Optimal learning rate schedules for SGD in changing data distributions.
problem Minimizing regret in online learning with changing data distributions.
method Characterized optimal schedules for linear regression, proposed schedules for general convex and non-convex losses, and defined a notion of regret for non-convex losses.
result Upper and lower bounds for regret with constants for convex losses, and an upper bound on total expected regret for non-convex losses.
New method detects changes by maximizing cross-entropy, outperforming existing techniques.
problem Detecting abrupt changes in data streams without labeled examples.
method Maximizes cross-entropy between segments to find change points, using dynamic programming.
result Outperforms three state-of-the-art approaches on challenging datasets.
Paper proposes algorithms to minimize both dynamic and adaptive regret simultaneously.
problem Traditional regret minimization algorithms are suboptimal for changing environments.
method Developed novel online algorithms to minimize dynamic and adaptive regret simultaneously.
result Proposed algorithms minimize dynamic and adaptive regret over any interval.
Body-worn video (BWV) cameras are increasingly utilized by police departments to provide a record of police-public interactions. However, large-scale BWV deployment produces terabytes of data per week, necessitating the development of effective computational methods to identify salient changes in video. In work carried…
We study exponential Levy models with change-point which is a random variable, independent from initial Levy processes. On canonical space with initially enlarged filtration we describe all equivalent martingale measures for change-point model and we give the conditions for the existence of f-divergence minimal equival…
Optimal transport between Gaussian Mixture Models improves domain adaptation efficiency.
problem Adapting machine learning models to new data distributions with minimal access.
method Optimal transport between Gaussian Mixture Models (GMMs) for domain adaptation.
result Our methods are more efficient and scalable with sample size and dimensions.
Domain adaptation framework identifies latent variables for target distribution identifiability.
problem Unsupervised domain adaptation without identifiable joint distribution of features and labels.
method Formulated latent variable model with invariant and changing components, constrained domain shift to influence only changing components.
result Joint distribution of data and labels in target domain is identifiable under mild conditions.
New method speeds up change-point detection in data sequences.
problem Efficiently detecting change-points in long data sequences.
method Sequential Gradient Descent and Quasi-Newton's Method.
result New method can be orders of magnitude faster than existing methods.
COMMOD debiases models with minimal and interpretable changes.
problem Inconsistent and costly model updates in fair machine learning.
method Introduced COMMOD, a novel algorithm for algorithmic fairness that minimizes changes and makes them interpretable.
result COMMOD achieves comparable performance to state-of-the-art debiasing methods while making minimal and interpretable changes.
KCUSUM detects abrupt changes in real-time data streams efficiently.
problem Detecting abrupt changes in high-volume scientific data streams.
method Kernel-based Cumulative Sum (KCUSUM) algorithm using Maximum Mean Discrepancy (MMD).
result KCUSUM outperforms traditional CUSUM in online change point detection.
A new buffer system improves continual learning in RL agents by adapting to changing environments.
problem Improving RL agents' ability to learn from changing environments over time.
method Multi-timescale replay buffer combined with invariant risk minimization.
result The method shows improvement over baselines in continual learning settings.
New algorithm reduces online learning regret by exploiting historical invariances.
problem Stochastic non-stationary linear bandits with changing reward models.
method ISD-linUCB algorithm that learns invariances in reward model.
result Significant regret improvements in fast-changing environments with historical data.
Efficient online kernel CUSUM detects changes quickly and accurately.
problem Detecting changes in online data streams efficiently.
method Online kernel CUSUM using maximum kernel statistics.
result Increased sensitivity to small changes compared to existing methods.
Let (N,g0) be a Kahler-Einstein surface with the first Chern class negative and assume that there exists a branched Lagrangian minimal surfaces with respect to the metric g0. We show that when the Kahler-Einstein metric is changed in the same component (i.e. the complex structure is changed), the Lagrangian m…
We show that the following unlinking strategy does not always yield an optimal sequence of crossing changes: first split the link with the minimal number of crossing changes, and then unknot the resulting components.
Study on blow-up behavior of sign-changing solutions for Yamabe equation.
problem Blow-up behavior of sign-changing solutions for Yamabe equation.
method Construction of a smooth metric on space forms to prove blow-up at lowest energy level.
result Blow-up occurs at the lowest energy level for sign-changing solutions in dimensions 11 to 24.
The study finds that most minimal surfaces in generic 4D manifolds intersect in complex ways.
problem Understanding self-intersections of minimal surfaces in generic Riemannian manifolds.
method Analyzing the properties of minimal surfaces in a generic Riemannian manifold of dimension four.
result Most minimal surfaces in generic 4D manifolds intersect in complex ways, with tangent planes failing to be complex with respect to any orthogonal complex structure.
Geometric pruning rules improve change point detection in multiple time series.
problem Detecting multiple changes in multiple independent time series.
method Dynamic programming algorithms with inequality-based and geometric pruning rules.
result Geometric pruning rules offer close-to-linear time complexity for multiple independent time series.
Study on abnormal curves in sub-Riemannian manifolds, proving length-minimizing properties.
problem Characterizing abnormal geodesics in sub-Riemannian manifolds.
method Analyzing curves that annihilate Lie brackets and proving minimization properties.
result Strictly abnormal geodesics can cease to be locally length-minimizing.
New algorithms detect and react to multiple change points in online learning.
problem Learning under multiple change points in environments with unknown and frequent shifts.
method Proposed Anytime Tracking CUSUM (ATC) algorithms that balance detection of significant shifts.
result Properly tuned ATC algorithms achieve nearly minimax-optimal performance.
Proves intersection properties of minimal hypersurfaces in various spaces.
problem Intersection properties of minimal hypersurfaces in different geometric settings.
method Two approaches: classifications of stable minimal hypersurfaces and conformal change with comparison geometry.
result Intersection properties for minimal hypersurfaces in specific geometric settings, including free boundary minimal hypersurfaces.
This study explains how different training methods affect the minimizer of neural networks.
problem How training methods influence the minimizer of neural networks.
method Explains how initialization size, adaptive optimization (AdaGrad), and stochastic mini-batch training affect the minimizer.
result Different training methods lead to different minimizers, even in overparameterized networks.
CROC identifies the earliest-changing stream as the root cause in multi-stream data.
problem Distribution-free root cause analysis in multi-stream data with unknown distributional changes.
method Conformal p-values and finite-sample valid confidence sets.
result CROC efficiently isolates the root cause under minimal assumptions.
The study connects lamination and orbit closures in hyperbolic manifolds.
problem Understanding the geometric and dynamical properties of horocycle orbit closures in Z-covers of compact hyperbolic manifolds. method Exposes connections between distance minimizing laminations and horospherical orbit closures in Z-covers of compact hyperbolic manifolds. Provides novel constructions and explicit descriptions. result Even slight perturbations to hyperbolic metrics can drastically change horocycle orbit closures.
New method detects market liquidity changes using order book data.
problem Detecting changes in market liquidity.
method Marked Hawkes processes and minimax quickest detection problem for doubly-stochastic Poisson process.
result Optimal stopping rule for detecting intensity changes in market liquidity.
We define a geometric flow that is designed to change surfaces of cylindrical type spanning two disjoint boundary curves into solutions of the Douglas-Plateau problem of finding minimal surfaces with given boundary curves. We prove that also in this new setting and for arbitrary initial data, solutions of the Teichmüll…
Change detection (CD) in time series data is a critical problem as it reveal changes in the underlying generative processes driving the time series. Despite having received significant attention, one important unexplored aspect is how to efficiently utilize additional correlated information to improve the detection and…
Two methods find at least two solutions to Kazdan-Warner's problem on surfaces.
problem Finding solutions to Kazdan-Warner's problem on two-dimensional surfaces.
method Direct method on convex sets and variational method of mountain pass.
result At least two solutions to the Kazdan-Warner's problem are found.
Proposes a method to repair arbitrage in option prices data.
problem Arbitrage in option price data can lead to poor performance or failure of financial applications.
method Formulates data repair as a linear programming (LP) problem to minimise price changes within bid and ask price bounds.
result The proposed method gives sparse perturbations on data and improves model calibration with enhanced robustness and reduced calibration error.
Develops efficient method for updating models with small data changes.
problem Efficiently updating models when data changes (e.g., adding/removing instances/features).
method Generalized Low-Rank Update (GLRU) for non-linear estimators.
result Provides updated solutions with computational complexity proportional to dataset changes.
This paper finds all prime alternating knots with minimal warping degree two.
problem Finding knots with minimal warping degree.
method Examined all prime alternating knots and determined those with minimal warping degree two.
result All prime alternating knots with minimal warping degree two were identified.
We propose algorithms for online principal component analysis (PCA) and variance minimization for adaptive settings. Previous literature has focused on upper bounding the static adversarial regret, whose comparator is the optimal fixed action in hindsight. However, static regret is not an appropriate metric when the un…
The paper explores how word embeddings affect the stability of downstream NLP models.
problem Small changes in training data can cause significant changes in model predictions.
method Empirical and theoretical analysis of embedding instability, including the introduction of eigenspace instability measure.
result Increasing embedding memory can reduce the disagreement in predictions by 5% to 37%.
Optimal search for change point anomaly in multiple processes.
problem Detecting a change point in an anomalous process among multiple normal processes.
method Deterministic search algorithm balancing sample complexity and detection accuracy.
result Asymptotically optimal in minimizing Bayes risk.
We consider the problem of distortion minimal morphing of n-dimensional compact connected oriented smooth manifolds without boundary embedded in Rn+1. Distortion involves bending and stretching. In this paper, minimal distortion (with respect to stretching) is defined as the infinitesimal relative change in vol…
Assuming minimal regularity assumptions on the data, we revisit the classical problem of finding isometric immersions into the Minkowski spacetime for hypersurfaces of a Lorentzian manifold. Our approach encompasses metrics having Sobolev regularity and Riemann curvature defined in the distributional sense, only. It ap…
In this work, we propose new objective functions to train deep neural network based density ratio estimators and apply it to a change point detection problem. Existing methods use linear combinations of kernels to approximate the density ratio function by solving a convex constrained minimization problem. Approximating…
To deal with changing environments, a new performance measure -- adaptive regret, defined as the maximum static regret over any interval, was proposed in online learning. Under the setting of online convex optimization, several algorithms have been successfully developed to minimize the adaptive regret. However, existi…
LaMBO optimizes modular systems with switching costs, achieving better results than existing methods.
problem Optimizing systems with costly variable updates in a sequence of modules.
method Lazy Modular Bayesian Optimization (LaMBO) that minimizes switching costs.
result LaMBO achieves vanishing regret and improves over existing cost-aware Bayesian optimization algorithms.