The paper introduces a method to measure the benefits of incidental supervision signals.
problem Lack of a principled way to measure the benefits of various types of incidental supervision signals.
method Unified PAC-Bayesian motivated informativeness measure, PABI.
result Demonstrates PABI's effectiveness in quantifying the value added by various types of incidental signals.
Paper tackles supervision bottleneck in machine learning.
problem Difficulty in generating supervision signals for learning models.
method Describes several learning paradigms to alleviate the supervision bottleneck.
result Illustrates the benefit of these paradigms in inducing semantic representations from text.
Big Data bring new opportunities to modern society and challenges to data scientists. On one hand, Big Data hold great promises for discovering subtle population patterns and heterogeneities that are not possible with small-scale data. On the other hand, the massive sample size and high dimensionality of Big Data intro…
We define a metric filtration of the Gordian graph by an infinite family of 1-dense subgraphs. The n-th subgraph of this family is generated by all knots whose fundamental groups surject to a symmetric group with parameter at least n, where all meridians are mapped to transpositions. Incidentally, we verify the Meridio…
Uniform covers with a finite-dimensional nerve are rare (i.e., do not form a cofinal family) in many separable metric spaces of interest. To get hold on uniform homotopy properties of these spaces, a reasonably behaved notion of an infinite-dimensional metric polyhedron is needed; a specific list of desired properties …
Classifies 7D manifolds with specific geometric properties.
problem Classifying 7D manifolds with parallel skew-symmetric torsion and G2 holonomy. method Extending Friedrich's work, using classification techniques for naturally reductive spaces and nearly parallel G2-structures. result Complete classification of 7D manifolds with the specified properties.
We consider the Hele-Shaw flow that arises from injection of two-dimensional fluid into a point of a curved surface. The resulting fluid domains have and are more or less determined implicitly by a mean value property for harmonic functions. We improve on the results of Hedenmalm and Shimorin \cite{HS} and obtain essen…
In this paper we investigate the asymptotics of forward-start options and the forward implied volatility smile in the Heston model as the maturity approaches zero. We prove that the forward smile for out-of-the-money options explodes and compute a closed-form high-order expansion detailing the rate of the explosion. Fu…
New interpretation of OT regularization as adversarial ground cost.
problem Optimal transport (OT) regularization for machine learning.
method Using Fenchel duality, interpret convex OT regularization as adversarial ground cost.
result Access to a robust dissimilarity measure on the ground space.
We introduce Parseval networks, a form of deep neural networks in which the Lipschitz constant of linear, convolutional and aggregation layers is constrained to be smaller than 1. Parseval networks are empirically and theoretically motivated by an analysis of the robustness of the predictions made by deep neural networ…
We present a theoretically grounded approach to train deep neural networks, including recurrent networks, subject to class-dependent label noise. We propose two procedures for loss correction that are agnostic to both application domain and network architecture. They simply amount to at most a matrix inversion and mult…
New pricing theory solves St. Petersburg paradox.
problem St. Petersburg Paradox unresolved for 280 years.
method Proposes new pricing theory with fair pricing rules.
result New pricing theory resolves the paradox.
We study a problem of the geometric quantization for the quaternion projective space. First we explain a Kaehler structure on the punctured cotangent bundle of the quaternion projective space, whose Kaehler form coincides with the natural symplectic form on the cotangent bundle and show that the canonical line bundle o…
Randomly guessing weights helps analyze RL benchmarks objectively.
problem Understanding the complexity of reinforcement learning benchmarks.
method Generate policy networks by randomly guessing their parameters, evaluate on benchmarks, and analyze results.
result Small untrained networks can provide a robust baseline for various RL tasks.
Researchers find optimal configurations of complex knots and links.
problem Finding the most efficient configurations of complex knots and links.
method Minimizing Möbius and Minimum Distance energies by describing them with a small number of free parameters.
result Optimal geometries for Hopf links, Borromean rings, and chain links are found.
Paper stabilizes bandit learning with regularization, improving inference under adaptive sampling.
problem Challenges in statistical inference with adaptive sampling.
method Refined stability condition for online algorithms, using regularized stochastic-mirror-descent-style methods.
result Derives precise regret bounds and asymptotic normality, showing necessity of regularization for valid inference.
This paper studies the infinitesimal structure of Carnot manifolds. By a Carnot manifold we mean a manifold together with a subbundle filtration of its tangent bundle which is compatible with the Lie bracket of vector fields. We introduce a notion of differential, called Carnot differential, for Carnot manifolds maps (…
Deep neural networks map brain lesions to deficits for better brain function understanding.
problem Mapping the functional brain organization from pathological lesions.
method Deep generative neural network architectures, specifically variational convolutional volumetric auto-encoders.
result Our model outperforms established methods in lesion-deficit inference across various scenarios.
A privacy-preserving synthetic data generation framework that distinguishes between true and phantom data disclosures.
problem Detecting and explaining data disclosures in synthetic datasets.
method Customizable empirical auditing framework with statistical hypothesis testing.
result Demonstrated tighter privacy leakage bounds than prior methods.
New defence against data-poisoning attacks in neural networks.
problem Data-poisoning attacks can evade existing defences and increase model efficacy.
method Proved geometric mechanism and identified near clone regime in input space.
result Regularisation and data augmentation reduce data fitting capacity and prevent poisoning.
The paper tackles robust statistical methods using Wasserstein DRO formulations.
problem Distributional uncertainty in learning from limited samples.
method Min-max distributionally robust optimization with Wasserstein DRO formulations.
result Error bounds free from the curse of dimensionality.
New DL algorithm detects critical chest X-ray findings without manual annotations.
problem Lack of explainability and manual annotation costs for DL models in medical imaging.
method Multi-instance learning approach to jointly classify and localize critical findings in CXR.
result Competitive classification results on three CXR datasets.
Paper develops inference methods for low-rank tensors without debiasing.
problem Statistical inference for low-rank tensor models.
method Two-iteration alternating minimization for asymptotic distribution.
result Asymptotic distributions and confidence regions for singular subspaces.
SupSiam and SupBYOL improve supervised representation learning with ANCL.
problem Improving supervised representation learning with ANCL.
method Proposed supervised ANCL framework leveraging labels to avoid collapse.
result Supervised ANCL improves representation learning across various datasets and tasks.
S4 learns new self-supervision automatically, improving accuracy with less human effort.
problem Lack of direct supervision in machine learning.
method Combines deep learning and probabilistic logic to automatically generate and verify new self-supervision.
result S4 can automatically propose accurate self-supervision, matching supervised methods with less human effort.
Paper analyzes if GANs with adversarial features outperform standard supervised learning.
problem Whether adversarial features improve performance over standard supervised learning.
method Theoretical analysis and empirical risk comparison.
result Supervised learning with adversarial features can outperform sole supervision under certain conditions.
Improves label propagation for weakly supervised learning.
problem Reducing the need for labeled data in machine learning.
method Label Propagation with Weak Supervision (LPA) analysis.
result Demonstrated improvements over existing methods on weakly supervised classification tasks.
Paper proposes an algorithm to recover full supervision from weakly labeled data.
problem Machine learning requires expensive data annotation, motivating the use of weak supervision.
method The paper introduces a disambiguation principle and an empirical disambiguation algorithm for partial labelling.
result The algorithm achieves exponential convergence rates under learnability assumptions.
Self-supervision provides effective representations for downstream tasks without requiring labels. However, existing approaches lag behind fully supervised training and are often not thought beneficial beyond obviating or reducing the need for annotations. We find that self-supervision can benefit robustness in a varie…
In this study, importance of user inputs is studied in the context of personalizing human activity recognition models using incremental learning. Inertial sensor data from three body positions are used, and the classification is based on Learn++ ensemble method. Three different approaches to update models are compared:…
Recent advances in semi-supervised learning have shown tremendous potential in overcoming a major barrier to the success of modern machine learning algorithms: access to vast amounts of human-labeled training data. Previous algorithms based on consistency regularization can harness the abundance of unlabeled data to pr…
The scarcity of data annotated at the desired level of granularity is a recurring issue in many applications. Significant amounts of effort have been devoted to developing weakly supervised methods tailored to each individual setting, which are often carefully designed to take advantage of the particular properties of …
Traditionally, there are three species of classification: unsupervised, supervised, and semi-supervised. Supervised and semi-supervised classification differ by whether or not weight is given to unlabelled observations in the classification procedure. In unsupervised classification, or clustering, all observations are …
Proposes a constrained labeling method for weakly supervised learning.
problem Combining weak supervision signals while navigating misleading correlations.
method Randomized constrained labeling within a defined space.
result Randomized constrained labeling converges after few iterations and outperforms other methods.
Self-supervised and supervised methods learn similar intermediate visual representations but diverge in final layers.
problem Comparing self-supervised and supervised methods for visual learning.
method Comparison of contrastive self-supervised and supervised methods on simple image data.
result Contrastive and supervised methods learn similar intermediate representations but diverge in final layers.
Improved VAE learns disentangled representations with less supervision.
problem Learning disentangled representations is challenging.
method Semi-supervised disentanglement learning with label replacement.
result Significant improvement in disentanglement with minimal supervision.
Paper develops methods for semi-supervised Fréchet regression.
problem High costs of obtaining non-Euclidean labels.
method Proposes semi-supervised NW Fréchet regression and semi-supervised kNN Fréchet regression.
result Demonstrates superior performance over supervised methods.
SPO optimizes LLMs by eliminating group-based baselines and variance issues.
problem Frequent degenerate groups and synchronization barriers in group-based policy optimization methods.
method Single-stream Policy Optimization (SPO) replaces per-group baselines with a persistent, KL-adaptive value tracker and global advantage normalization.
result SPO converges more smoothly and attains higher accuracy than GRPO, improving maj@32 by +3.4 pp across five math benchmarks.
Paper resolves the debate on process vs. outcome supervision in reinforcement learning.
problem Distinguishing between process and outcome supervision in reinforcement learning.
method Developed a technical tool (Change of Trajectory Measure Lemma) to show equivalence between outcome and process supervision under standard data coverage assumptions.
result Reinforcement learning through outcome supervision is statistically equivalent to process supervision, up to polynomial factors in horizon.
Optimal and safe semi-supervised learning estimator for high-dimensional data.
problem Improving regression parameter estimation with unlabeled data in high-dimensional settings.
method Established minimax lower bound, proposed optimal and safe semi-supervised estimators.
result Optimal semi-supervised estimator achieves the minimax lower bound.
Survey of self-supervised learning methods in computer vision, NLP, and graph learning.
problem Manual labeling and vulnerability to attacks in supervised learning.
method Generative, contrastive, and generative-contrastive approaches.
result Self-supervised learning achieves high performance in representation learning.
Improved self-supervised learning for document images.
problem Performance of self-supervised pre-training on document images is poor.
method Proposed context-aware alternatives and a novel multi-modal method.
result Novel method outperforms other self-supervised methods on document image classification.
Weak supervision challenges black-box models, suggesting fusion of modeling cultures.
problem Challenges of strong supervision in achieving accurate predictions.
method Integrating data modeling into algorithmic modeling for weak supervision.
result Integration of data modeling culture improves model stability and accuracy.
Proposes a framework for semi-supervised continual learning from sequentially arriving data.
problem Learning from data with changing task distribution over time, especially in domains with a mix of labeled and unlabeled data.
method Meta-Consolidation for Continual Semi-Supervised Learning (MCSSL) framework with a hypernetwork and semi-supervised auxiliary classifier.
result Significant improvements in continual semi-supervised learning setting.
Unified framework for semi-supervised learning reduces annotation needs.
problem Sparse annotations and large amounts of unlabeled data in computational pathology.
method S5CL integrates fully-supervised, self-supervised, and semi-supervised learning through hierarchical contrastive losses.
result S5CL improves accuracy and F1-score in histopathological datasets with sparse labels.
Self-supervision improves GCNs' generalizability and robustness.
problem Improving graph convolutional networks' performance.
method Three mechanisms of self-supervision, multi-task learning, and graph adversarial training.
result Self-supervision enhances GCNs' robustness and generalizability.
We present a technique to improve the transferability of deep representations learned on small labeled datasets by introducing self-supervised tasks as auxiliary loss functions. While recent approaches for self-supervised learning have shown the benefits of training on large unlabeled datasets, we find improvements in …
For semi-supervised techniques to be applied safely in practice we at least want methods to outperform their supervised counterparts. We study this question for classification using the well-known quadratic surrogate loss function. Using a projection of the supervised estimate onto a set of constraints imposed by the u…