New method improves model risk prediction using cross-audit projection.
problem Over-optimism in K-fold CV for binary classification. method Cross-audit projection (CAP) procedure combining resampling and asymptotic bias correction.
result CAP estimator achieves second-order asymptotic unbiasedness.
Proposes DRRO to mitigate over-optimization in RLHF from human feedback.
problem Over-optimization due to reward misspecification in RLHF.
method Wasserstein distributionally robust regret optimization (DRRO).
result DRRO mitigates over-optimization more effectively than existing baselines.
Detecting adversarial examples currently stands as one of the biggest challenges in the field of deep learning. Adversarial attacks, which produce adversarial examples, increase the prediction likelihood of a target class for a particular data point. During this process, the adversarial example can be further optimized…
In a context where most published articles are devoted to the development of "new methods", comparison studies are generally appreciated by readers but surprisingly given poor consideration by many scientific journals. In connection with recent articles on over-optimism and epistemology published in Bioinformatics, thi…
The paper tackles Neyman-Pearson classification control issues.
problem Neyman-Pearson classification's control constraint is hard to satisfy in finite samples.
method Developed refined learning procedures under two accuracy control strategies.
result Proposed methods achieve desired control levels in finite samples.
The paper investigates Goodhart's law and its impact on goal alignment.
problem The adverse effects of optimizing a measure when it diverges from the true goal.
method Formal analysis of Goodhart's law, focusing on the tail distribution of discrepancies.
result Goodhart's law depends on the tail distribution of discrepancies between the true goal and the optimized measure.
This paper develops a model of reference-dependent assessment of subjective beliefs in which loss-averse people optimally choose the expectation as the reference point to balance the current felicity from the optimistic anticipation and the future disappointment from the realisation. The choice of over-optimism or over…
Study wSAA for contextual decisions, improving uncertainty quantification under computational constraints.
problem Uncertainty quantification limitations in wSAA for contextual stochastic optimization.
method Establish central limit theorems and asymptotic-normality-based confidence intervals for optimal costs.
result Over-optimizing can mitigate misspecification and preserve asymptotic normality, albeit at a slower convergence rate.
The computational costs of inference and planning have confined Bayesian model-based reinforcement learning to one of two dismal fates: powerful Bayes-adaptive planning but only for simplistic models, or powerful, Bayesian non-parametric models but using simple, myopic planning strategies such as Thompson sampling. We …
The study evaluates how prediction helps identify the worst-off in welfare programs.
problem Identifying the most vulnerable individuals for support.
method Mathematical models and real-world case study on long-term unemployment.
result Prediction is more effective than other policy levers in surfacing the worst-off.
Optimized parallel algorithms for identifying strong ties in data.
problem Identifying strong ties in data with varying distances and community sizes.
method Design and analysis of sequential and parallel algorithms for partitioned local depths.
result Optimized algorithms achieve up to 19.4x speedup in parallel execution.
Unified parametric assumption improves convergence guarantees for nonconvex optimization.
problem Weak convergence guarantees for nonconvex optimization.
method Introducing a novel unified parametric assumption.
result Unified convergence theorem for gradient-based methods.
Model-free Reinforcement Learning (RL) works well when experience can be collected cheaply and model-based RL is effective when system dynamics can be modeled accurately. However, both assumptions can be violated in real world problems such as robotics, where querying the system can be expensive and real-world dynamics…
Delayed rejection HMC improves sampling efficiency for multiscale distributions.
problem Hamiltonian Monte Carlo struggles with wide-ranging distributions, especially in high-curvature areas.
method Introduces a delayed rejection variant of HMC, using geometrically smaller step sizes for retries.
result Up to five-fold performance gains in effective sample size per gradient evaluation.
Nesterov SGD is widely used for training modern neural networks and other machine learning models. Yet, its advantages over SGD have not been theoretically clarified. Indeed, as we show in our paper, both theoretically and empirically, Nesterov SGD with any parameter selection does not in general provide acceleration o…
Optimism stabilizes Thompson Sampling for adaptive inference in multi-armed bandits.
problem Subtle inferential properties of Thompson Sampling under adaptive data collection.
method Introduced optimism as a key mechanism to restore stability and validity of inference.
result Suitably implemented optimism stabilizes Thompson Sampling and enables asymptotically valid Wald inference.
Study develops ensemble machine learning framework for predicting groundwater heavy metal pollution.
problem Statistical complexity and spatial heterogeneity of heavy metal contamination in groundwater.
method Nested cross-validated ensemble machine learning with response transformations (raw, log, Gaussian copula).
result Copula-based models with DBSCAN clustering diagnostics provide the most reliable and interpretable assessments of groundwater contamination.
This paper investigates teacher hacking during language model distillation and proposes methods to mitigate it.
problem Teacher hacking during language model distillation, leading to suboptimal performance.
method A controlled experimental setup involving an oracle LM, teacher LM, and student LM, using fixed offline or online data generation techniques.
result Data diversity is the key factor in preventing teacher hacking during distillation.
QAlign improves language model alignment with less compute, outperforming existing methods.
problem Improving language model performance with limited test-time computation.
method QAlign: sampling from optimal aligned distribution using Markov chain Monte Carlo.
result Consistent improvements over existing methods on various benchmarks.
Optimal liquidation strategy reduces risk and improves performance.
problem Risk-averse trader liquidating assets in volatile markets.
method Conditional Value-at-Risk (CVaR) objective, dynamic trading strategies, closed-form solutions.
result Dynamic policies outperform static and VWAP policies by 15-25%.