Paper improves text-to-SQL models with schema-aware denoising.
problem Text-to-SQL models struggle with schema linking and grammar correctness.
method Adapts transformer-based seq-to-seq model with SeaD denoising objectives and clause-sensitive decoding.
result Improves seq-to-seq model performance on WikiSQL benchmark.
A new method for uplift modeling using learning-to-rank techniques.
problem Improving customer targeting in marketing and retention.
method Unified formalization of uplift measures, learning-to-rank with PCG metric, LambdaMART optimization.
result Improved results compared to standard learning-to-rank metrics and state-of-the-art uplift modeling.
Symmetries of Poisson manifolds are in general quantized just to symmetries up to homotopy of the quantized algebra of functions. It is therefore interesting to study symmetries up to homotopy of Poisson manifolds. We notice that they are equivalent to Poisson principal bundles and describe their quantization to symmet…
This paper proposes learning to jump for generative modeling of sparse, skewed, heavy-tailed data.
problem Limited ability of diffusion models in modeling sparse, skewed, heavy-tailed data.
method Forward count thinning process and reverse count thickening process to train a deep neural network.
result Learning to jump performs better than learning to denoise for non-negative, sparse data.
Recurrent Neural Networks (RNNs) are powerful models for sequential data that have the potential to learn long-term dependencies. However, they are computationally expensive to train and difficult to parallelize. Recent work has shown that normalizing intermediate representations of neural networks can significantly im…
Deep RL methods improve resource allocation in uncertain environments.
problem Optimizing resource allocation in dynamic, uncertain environments.
method Developed three DDPG-based approaches to handle constraints and combinatorial action spaces.
result Demonstrated improved performance over existing methods on real and semi-real data.
This belongs to a series of papers devoted to the study of the cohomology of classifying spaces of Lie groupoids. Our aim here is to introduce and study the notion of representation up to homotopy of Lie groupoids, the resulting derived category, and to show that the adjoint representation is well defined as a represen…
This paper uses deep reinforcement learning to generate profitable trading signals in financial markets.
problem Generating consistent profitable trading signals in any financial market.
method Developed a novel Markov decision process (MDP) model to capture financial market dynamics and used deep reinforcement learning to learn profitable trading strategies.
result The model generates positively robust performance in two different financial markets.
Improves text-to-SQL models by selecting the best SQL query from beam output.
problem Simplifying database query writing for natural language questions.
method Discriminative re-ranker using BERT fine-tuned classifier.
result Achieved top 4 score on Spider leaderboard.
Proposes isotonic recalibration for insurance pricing to ensure auto-calibration under low signal-to-noise ratio.
problem Ensuring auto-calibration in insurance pricing models to prevent cross-financing.
method Applies isotonic recalibration to regression models to achieve auto-calibration.
result Isotonically recalibrated regression functions have low complexity under low signal-to-noise ratio.
Meta-learn causal structures based on adaptation speed to sparse distributional changes.
problem Learning cause-effect relationships from sparse distributional changes.
method Meta-learn causal structures based on adaptation speed to modified distributions.
result Sparse expected gradients and lower effective degrees of freedom lead to faster adaptation.
Proposes vMF distribution for skewed elliptical distributions.
problem Skewed distributions not adequately modeled by symmetric distributions.
method Introduces von-Mises-Fisher (vMF) distribution to represent skewed elliptical distributions.
result vMF distribution provides an explicit and simple probability representation of skewed elliptical distributions.
Proposes a machine learning framework for more efficient economic dispatch.
problem Temporal and spatial correlations between system cost and load prediction errors.
method End-to-end machine learning approach with task-specific learning criteria and an efficient optimization kernel.
result Demonstrates the effectiveness and efficiency of the proposed learning framework.
Model shows how banks' hidden-to-maturity accounting can mask run risk and lead to financial instability.
problem Run risk and hidden-to-maturity accounting in banking systems.
method Balance sheet model and optimization problem to assess run risk and resilience.
result Held-to-maturity accounting can mask revaluation losses and increase run risk.
Proposes minimal interventions over counterfactual explanations for algorithmic recourse.
problem Lack of actionable recommendations for algorithmic recourse.
method Causal reasoning to shift focus from explanations to recommendations.
result Minimal interventions provide more actionable recommendations for recourse.
BPE improves text-to-SQL generation by reducing training time and improving accuracy.
problem Improving text-to-SQL generation accuracy with neural models.
method Adapted Byte-Pair Encoding (BPE) for SQL generation, introduced a novel stopping criterion, and used AST BPE.
result Improved accuracy on 5 out of 6 English text-to-SQL tasks, reduced training time by 50%.
The paper finds obstructions to Lie algebroid representations up to homotopy.
problem Obstacles to representations up to homotopy of Lie algebroids.
method Analyzes Pontryagin characters and vector bundles over Lie algebroids.
result Vanishing of Pontryagin characters implies no representation up to homotopy.
Invites readers to explore higher Teichmüller theory.
problem None explicitly stated; focuses on introduction.
method Description and overview of facets.
result Introduces higher Teichmüller theory to readers.
Ebay uses forecasting and simulation to decide when to disable a vendor.
problem Determining the optimal time to disable a vendor to avoid customer loss.
method Data-driven approach involving multiplicative seasonality model, Monte Carlo simulation, and linear model.
result Identifies the best time to disable a vendor to minimize customer loss.
A new method to dynamically manage dark liquidity to balance capturing available liquidity and protecting from signalling.
problem Limiting access to dark venues and imposing minimum fill sizes restricts liquidity and can lead to excessive signalling.
method Dynamic monitoring of dark liquidity on a per fill basis, allowing real-time adjustments to trading exposure.
result The method allows traders to maximize available liquidity while protecting from excessive signalling.
End-to-end autonomous driving models get better uncertainty estimates.
problem Uncertainty quantification for end-to-end autonomous driving models.
method Approximate inference for implicit copula neural linear model.
result Densities for steering angle are marginally calibrated.
The paper introduces invariants to describe period-doubling routes to chaos in dynamical systems.
problem Understanding the dynamics of period-doubling routes to chaos in complex systems.
method Introducing three topological invariants to describe the topology of period-doubling routes to chaos.
result Ascribed symbolic dynamics to perturbations of the Shilnikov homoclinic scenario and dynamics of the Henon map.
Noise injection improves inference privacy in DNN models.
problem Malicious servers can infer sensitive attributes from input data.
method Adaptive Noise Injection (ANI) using a lightweight DNN on the client.
result Significant improvement in privacy (up to 48.5% degradation in sensitive-task accuracy with <1% degradation in primary accuracy).
Proposes a method to prevent forgetting in neural networks by training an update predictor.
problem Catastrophic forgetting in neural networks.
method Train an update predictor to minimize loss on past and current tasks.
result The proposed method effectively mitigates catastrophic forgetting in continual learning.
End-to-end learning may not scale well to complex tasks.
problem The scalability of end-to-end learning to complex and diverse data processing architectures.
method Simple experiments to demonstrate inefficiencies in end-to-end learning.
result End-to-end learning does not make optimal use of modular design and can lead to complete breakdown of learning.
Entropy regularization improves policy optimization in reinforcement learning.
problem Improving policy optimization in reinforcement learning.
method Entropy regularization is introduced to soften the greedy policy towards a more diverse softmax policy, leading to a continuously parameterized algorithm that interpolates between policy gradient and Q-learning.
result An intermediate algorithm can improve performance in reinforcement learning.
New attacks reprogram neural networks to perform new tasks.
problem Neural networks are vulnerable to adversarial attacks that can cause mistakes or specific outputs.
method Develops attacks that reprogram models to perform tasks chosen by the attacker, without specifying outputs.
result Demonstrates reprogramming on six models, including counting and classification tasks.
Extends image-to-image translation to multiple distributions, allowing composite functions.
problem Limited to single pair translations, new mechanism scalable to multiple distributions.
method Decoupled training mechanism for multiple distributions, composite translation functions.
result Generates images with characteristics not seen in training set.
Users can anticipate follower preferences by balancing feedback exploitation and exploration.
problem How users can anticipate their followers' preferences based on feedback.
method Theoretical analysis and practical algorithms for sequential decision making and utility maximization.
result Users need to balance exploitation and exploration to succeed in anticipating follower preferences.
A new algorithm uses reservoir sampling to enhance a reinforcement learning agent's memory.
problem Efficiently maintaining and recalling past states for reinforcement learning.
method Reservoir sampling to maintain a fixed number of past states for an external memory.
result The method allows for efficient online computation of gradient estimates.
Improves text-to-image translation by using GANs and captioning networks.
problem Generating images that accurately reflect the meaning of a sentence.
method Uses cycle consistent adversarial networks and captioning networks to improve image generation.
result Significantly improved image quality compared to existing methods.
Simplified method to list 250 knots with up to 10 crossings.
problem Generating a list of knots with up to 10 crossings.
method Generating all planar knot diagrams, simplifying, grouping, reducing with moves, and using invariants.
result Proves there are exactly 250 knots with up to 10 crossings.
Study examines perceptions and attitudes about breast cancer on Twitter.
problem Understanding public perceptions and attitudes towards breast cancer on social media.
method Identified and collected tweets, used topic modeling and sentiment analysis.
result Identified themes and quantified users' perceptions and emotions about breast cancer.
Study uses NLP to analyze emotions and challenges of young people with IDD.
problem Challenges faced by young people with IDD during transition to adulthood.
method Natural language processing, unsupervised machine learning, topic modeling.
result NLP methods can assist psychologists in analyzing emotions and summarizing key topics.
Homotopy theory applied to singular foliations leads to new results.
problem Existence and uniqueness of universal L∞-algebroids for singular foliations. method Applied homotopy theory to left semi-model categories and L∞-algebroids. result Recovery of results similar to Laurent-Gengoux and al. about universal L∞-algebroids. We consider the problem of fitting a linear model to data held by individuals who are concerned about their privacy. Incentivizing most players to truthfully report their data to the analyst constrains our design to mechanisms that provide a privacy guarantee to the participants; we use differential privacy to model in…
End-to-end training of DBMs with improved gradient estimation.
problem Biased gradient estimation in DBMs, especially with high-dimensional states.
method Unbiased contrastive divergence using MH coupling and local mode initialization.
result End-to-end training of DBMs without greedy pretraining, achieving FID score of 10.33 for MNIST.
Calibrating agent-based models in finance is challenging but crucial.
problem Difficulty in calibrating agent-based models due to computational expense.
method Applied an established financial agent-based model calibration framework to a simplified model of high- and low-frequency trader interaction.
result Demonstrated inadequacies of stylized fact-centric validation methods.
We observe that any regular Lie groupoid G over an manifold M fits into an extension K→G→E of a foliation groupoid E by a bundle of connected Lie groups K. If $\FF$ is the foliation on M given by the orbits of E and T is a complete transversal to $\FF$, this extension restricts to T, as an extension $K_{T}\to…
Paper explores MM strategies that can refuse to quote or provide single-sided quotes.
problem Overcoming risks in market making due to changing market conditions.
method Adversarial reinforcement learning with new MM agent designs.
result Refusal to quote or providing single-sided quotes can improve MM performance.
Graphs mapped to projective plane classified up to deformation.
problem Classifying immersions of graphs to the projective plane.
method Classification via regular homotopy and construction of invariants.
result Complete invariant for equivalence classes of immersions.
Proposes a method to quantify and explain deep learning model uncertainties.
problem Deep learning model predictions are sensitive to perturbations and adversarial attacks.
method Gradient-based uncertainty attribution method to identify problematic regions and propose mitigation strategies.
result Proposed UA-Backprop method achieves competitive accuracy and efficiency compared to existing methods.
Given a compact Riemannian manifold, with positive Yamabe quotient, not conformally diffeomorphic to the standard sphere, we prove a priori estimates for solutions to the Yamabe problem. We restrict ourselves to the dimensions less than or equal to 7, where the Positive Mass Theorem is known to be true. We also show th…
SequenceR uses seq-to-seq learning to fix bugs in code.
problem Fixing bugs in code using machine learning.
method Sequence-to-sequence learning with copy mechanism, trained on curated code samples.
result SequenceR can perfectly predict and find correct patches for bugs.
Bayesian optimisation generates saliency maps for black-box models.
problem Generating saliency maps for models without access to parameters.
method Bayesian optimisation sampling method to find global salient regions.
result Approach outperforms grid-based methods and performs similarly to gradient-based methods.
Study shows adversarial robustness and common perturbation robustness are independent.
problem Understanding the relationship between adversarial robustness and common perturbation robustness in neural networks.
method Conducted experiments to benchmark neural network robustness to common perturbations and adversarial examples.
result Adversarial robustness and common perturbation robustness are independent attributes.
Simulates DeLend Platform behavior to optimize operational parameters.
problem Optimizing the DeLend Platform's operational parameters.
method Agent-based simulations to connect and test different agent sets.
result Estimates how key variables respond to different policies.
Developed a new formalism to describe Riemannian geometries using geodesic flow bundles.
problem Understanding the consequences of Einstein equations without solving metric equations.
method Using the bundle of arclength parametrized geodesics (geodesic flow bundle GFB) to describe Riemannian geometry.
result Generalized the cosine- and sine-laws for constant curvature to varying curvature fields.