On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,695 papers · 148 categories

Trend · papers per month

count % of corpus

48 results for “Sequence Likelihood Calibration”

Temporal Probability Calibrationcs.LG

The paper addresses probability calibration for incomplete sequences.

problem Improving probability estimates from incomplete sequences.

method Adapting traditional calibration techniques to sequences of varying lengths.

result Proposed methods improve probability calibration for modern sequential models.

probability calibration incomplete sequences sequential data model calibration machine learning

2020-02-07abs ↗pdf ↗

Noise Titration: Exact Distributional Benchmarking for Probabilistic Time Series Forecastingcs.LG

Noise titration benchmarks time series forecasting models rigorously.

problem Evaluation of time series forecasting models is often flawed due to lack of interventionist methods.

method Interventionist benchmarking using Gaussian noise titration of dynamical systems.

result Fern model outperforms state-of-the-art models in non-stationary conditions.

time series forecasting benchmarking Gaussian noise chaotic systems stochastic dynamics

2026-03-23abs ↗pdf ↗

Calibration of Phone Likelihoods in Automatic Speech Recognitionstat.ML

In this paper we study the probabilistic properties of the posteriors in a speech recognition system that uses a deep neural network (DNN) for acoustic modeling. We do this by reducing Kaldi's DNN shared pdf-id posteriors to phone likelihoods, and using test set forced alignments to evaluate these using a calibration s…

2016-06-14abs ↗pdf ↗

Maximum Likelihood with Bias-Corrected Calibration is Hard-To-Beat at Label Shift Adaptationcs.LG

Label shift refers to the phenomenon where the prior class probability p(y) changes between the training and test distributions, while the conditional probability p(x|y) stays fixed. Label shift arises in settings like medical diagnosis, where a classifier trained to predict disease given symptoms must be adapted to sc…

2019-01-21abs ↗pdf ↗

Revisiting Explicit Regularization in Neural Networks for Well-Calibrated Predictive Uncertaintycs.LG

This work improves neural network calibration using explicit regularization.

problem Improving predictive uncertainty in neural networks.

method Introducing a probabilistic calibration measure and exploring explicit regularization techniques.

result Explicit regularization improves log-likelihood and predictive uncertainty.

neural networks calibration uncertainty explicit regularization log-likelihood

2020-06-11abs ↗pdf ↗

Unifying Simulation and Inference with Normalizing Flowshep-ph

Unified detector calibration and simulation using MLE from generative models.

problem Combining detector calibration and simulation using traditional methods.

method Maximum likelihood estimation from conditional generative models.

result Prior-independent and non-Gaussian resolutions possible.

deep learning detector calibration maximum likelihood generative models energy regression

2024-04-29abs ↗pdf ↗

PEP: Parameter Ensembling by Perturbationcs.LG

PEP improves deep network performance and calibration by perturbing optimal parameters.

problem Improving deep network performance and calibration.

method Parameter Ensembling by Perturbation (PEP) constructs an ensemble of parameter values as random perturbations of the optimal set, maximizing log-likelihood on validation data.

result PEP provides a small to substantial improvement in calibration and log-likelihood, and in some cases, classification accuracy.

deep learning ensemble parameter perturbation log-likelihood calibration

2020-10-24abs ↗pdf ↗

X-CAL: Explicit Calibration for Survival Analysiscs.LG

X-CAL improves survival model calibration without sacrificing predictive power.

problem Improving the calibration of survival models to better match observed data.

method Explicit calibration (X-CAL) turns distributional calibration into a differentiable objective for survival modeling.

result X-CAL improves calibration metrics without significantly reducing predictive performance.

survival analysis calibration differentiable objective maximum likelihood estimation mimic-iii

2021-01-13abs ↗pdf ↗

SMURF-THP: Score Matching-based UnceRtainty quantiFication for Transformer Hawkes Processcs.LG

SMURF-THP improves Transformer Hawkes process models by providing uncertainty quantification.

problem Uncertainty quantification for Transformer Hawkes process predictions.

method Score matching for learning the score function of event arrival times.

result SMURF-THP outperforms likelihood-based methods in confidence calibration.

Transformer Hawkes process uncertainty quantification score matching event sequence modeling confidence intervals

2023-10-25abs ↗pdf ↗

Calibration of the rating transition model for high and low default portfoliosq-fin.RM

The paper develops ML algorithms for calibrating credit rating transition models for high and low default portfolios.

problem Calibration of credit rating transition models for high and low default portfolios.

method Developed Maximum likelihood (ML) algorithms, including Laplace approximation for high-default portfolios and particle filter with Gaussian process regression for low-default portfolios.

result Both algorithms produce accurate approximations of the likelihood function and ML estimates of model parameters.

credit rating transition maximum likelihood Kalman filter particle filter Gaussian process regression

2024-05-01abs ↗pdf ↗

Online monotone density estimation and log-optimal calibrationstat.ML

Study online monotone density estimation with expert aggregation and log-optimal calibration.

problem Online monotone density estimation and log-optimal calibration.

method Proposed two online estimators: Grenander estimator and expert aggregation estimator.

result Online estimators achieve

O(n^{1/3})

cumulative log-likelihood gap and

\sqrt{n\log{n}}

pathwise regret bound.

online density estimation monotone density log-optimal calibration exponential weighting stochastic setting

2026-02-09abs ↗pdf ↗

Variational Variance: Simple, Reliable, Calibrated Heteroscedastic Noise Variance Parameterizationcs.LG

A simple method treats heteroscedastic variance variatively, improving model calibration and sample quality.

problem Brittle optimization impacts model likelihoods for mean and variance estimation.

method Proposes a variational approach to heteroscedastic variance, improving predictive mean and variance calibration.

result The proposed method significantly improves parameter calibration and sample quality for regression and VAEs.

heteroscedastic variance variational approach model calibration predictive distribution neural networks

2020-06-08abs ↗pdf ↗

An efficient likelihood-free Bayesian inference method based on sequential neural posterior estimationstat.ML

A new method improves SNPE for intractable likelihood models.

problem Simulation-based models with intractable likelihoods.

method Adaptive calibration kernel and variance reduction techniques.

result The proposed method provides a better approximation of the posterior.

SNPE likelihood-free inference Bayesian neural networks variance reduction

2023-11-21abs ↗pdf ↗

Human Alignment of Large Language Models through Online Preference Optimisationcs.LG

Paper shows equivalence between two alignment methods and introduces a new algorithm.

problem Ensuring human alignment of large language models for useful, safe, and pleasant user experience.

method Introduces IPO-MD algorithm, showing equivalence between IPO and Nash-MD methods.

result Equivalence between IPO and Nash-MD methods proven when considering online version of IPO.

Human Alignment Large Language Models Reinforcement Learning Policy Optimisation Sequence Likelihood Calibration

2024-03-13abs ↗pdf ↗

The art of probability-of-default curve calibrationq-fin.RM

PD curve calibration refers to the transformation of a set of rating grade level probabilities of default (PDs) to another average PD level that is determined by a change of the underlying portfolio-wide PD. This paper presents a framework that allows to explore a variety of calibration approaches and the conditions un…

2012-12-15abs ↗pdf ↗

Resultant: Incremental Effectiveness on Likelihood for Unsupervised Out-of-Distribution Detectionstat.ML

Resultant improves likelihood-based U-OOD detection across various tasks.

problem Improving likelihood-based U-OOD detection performance.

method Resultant combines post-hoc prior and dataset entropy-mutual calibration techniques.

result Resultant achieves new state-of-the-art U-OOD detection performance.

likelihood unsupervised out-of-distribution calibration post-hoc

2024-09-05abs ↗pdf ↗

On Calibrating Diffusion Probabilistic Modelscs.LG

New method calibrates DPMs to improve likelihood bounds.

problem Improving the likelihood bounds of DPMs.

method Deriving concentration bounds and using the optional stopping theorem for data scores to calibrate DPMs.

result Calibrated DPMs can increase likelihood bounds and improve sampling quality.

Diffusion Probabilistic Models Calibration Score Matching Martingale Generative Models

2023-02-21abs ↗pdf ↗

MLE-guided parameter search for task loss minimization in neural sequence modelingcs.LG

A new method optimizes neural sequence models for better task performance.

problem Training neural sequence models with maximum likelihood estimation ignores task losses.

method Maximum likelihood guided parameter search (MGS) in the parameter space.

result MGS optimizes sequence-level losses, reducing repetition and non-termination.

neural sequence modeling maximum likelihood estimation task loss parameter search random search

2020-06-04abs ↗pdf ↗

It Is Likely That Your Loss Should be a Likelihoodcs.LG

Optimizing full likelihoods adapts loss scales and shapes for robust modeling.

problem Rigid loss functions limit model adaptability and robustness.

method Optimize full likelihoods with adjustable parameters.

result Adaptive tuning of loss scales and shapes improves model robustness.

likelihood optimization robust modeling loss adaptability outlier detection re-calibration

2020-07-12abs ↗pdf ↗

General Bayesian Updating and the Loss-Likelihood Bootstrapstat.ME

In this paper we revisit the weighted likelihood bootstrap, a method that generates samples from an approximate Bayesian posterior of a parametric model. We show that the same method can be derived, without approximation, under a Bayesian nonparametric model with the parameter of interest defined as minimising an expec…

2017-09-22abs ↗pdf ↗

Bayesian calibration for forensic evidence reportingstat.ML

We introduce a Bayesian solution for the problem in forensic speaker recognition, where there may be very little background material for estimating score calibration parameters. We work within the Bayesian paradigm of evidence reporting and develop a principled probabilistic treatment of the problem, which results in a…

2014-03-24abs ↗pdf ↗

Bayesian model calibration for block copolymer self-assembly: Likelihood-free inference and expected information gain computation via measure transportphysics.comp-ph

Bayesian calibration for BCP self-assembly models using image data and measure transport.

problem Calibrating models of BCP self-assembly from image data with aleatory uncertainty.

method Likelihood-free inference via measure transport and summary statistics.

result Expected information gains can be computed efficiently for model calibration.

block copolymer self-assembly Bayesian calibration measure transport summary statistics

2022-06-22abs ↗pdf ↗

Unifying Likelihood-free Inference with Black-box Optimization and Beyondcs.LG

Unified approach for sequence design combining likelihood-free inference and black-box optimization.

problem Designing biological sequences efficiently and accurately.

method Unified probabilistic framework integrating likelihood-free inference and black-box optimization.

result Previous optimization methods can be adapted and new algorithms proposed within this framework.

likelihood-free inference black-box optimization sequence design probabilistic framework

2021-10-06abs ↗pdf ↗

Post-hoc Calibration of Neural Networks by g-Layerscs.LG

Post-hoc calibration of neural networks using g-Layers proves theoretical justification.

problem Ensuring the confidence of neural network decisions in real-world applications.

method Proves theoretical justification for post-hoc calibration methods by adding g-Layers and minimizing NLL.

result Proves that adding g-Layers and minimizing NLL can lead to a calibrated network.

Neural Networks Calibration Post-Hoc Calibration g-Layers Negative Log-Likelihood

2020-06-23abs ↗pdf ↗

Multi-Class Gaussian Process Classification Made Conjugate: Efficient Inference via Data Augmentationstat.ML

We propose a new scalable multi-class Gaussian process classification approach building on a novel modified softmax likelihood function. The new likelihood has two benefits: it leads to well-calibrated uncertainty estimates and allows for an efficient latent variable augmentation. The augmented model has the advantage …

2019-05-23abs ↗pdf ↗

Breaking the Likelihood Trap: Variance-Calibrated Modulation for Large Language Model Decodingcs.CL

Variance-Calibrated Modulation (VCM) addresses the likelihood trap in LLMs by reshaping the probability distribution before truncation.

problem LLMs fall into the likelihood trap, leading to repetitive degeneration and vocabulary dullness.

method VCM reshapes the probability distribution before truncation through Contextual Searchlight and Adaptive Self-Debiasing.

result VCM mitigates the likelihood trap across open-ended generation, factual QA, and mathematical reasoning.

LLM likelihood trap decomposition post-hoc training-free

2026-06-21abs ↗pdf ↗

Conformal Predictive Systems Under Covariate Shiftcs.LG

WCPS extends CPS to handle covariate shifts, providing probabilistically calibrated predictions.

problem Applying CPS to scenarios with covariate shifts.

method WCPS uses likelihood ratios between training and testing covariate distributions.

result WCPS are probabilistically calibrated under covariate shift.

CPS covariate shift WCPS likelihood ratios probabilistic calibration

2024-04-23abs ↗pdf ↗

Failure of Calibration is Typicalmath.ST

Schervish (1985b) showed that every forecasting system is noncalibrated for uncountably many data sequences that it might see. This result is strengthened here: from a topological point of view, failure of calibration is typical and calibration rare. Meanwhile, Bayesian forecasters are certain that they are calibrated-…

2013-06-20abs ↗pdf ↗

Calibration of Encoder Decoder Models for Neural Machine Translationcs.LG

We study the calibration of several state of the art neural machine translation(NMT) systems built on attention-based encoder-decoder models. For structured outputs like in NMT, calibration is important not just for reliable confidence with predictions, but also for proper functioning of beam-search inference. We show …

2019-03-03abs ↗pdf ↗

Exact Stiefel Optimization for Probabilistic PLS: Closed-Form Updates, Error Bounds, and Calibrated Uncertaintystat.ML

A new framework for PPLS combines noise estimation, optimization, and calibration.

problem Probabilistic PLS models need interpretable latent factors and calibrated uncertainty.

method End-to-end pipeline combining noise estimation, constrained optimization, and prediction calibration.

result Achieves near-nominal coverage and native calibrated uncertainty across benchmarks.

Probabilistic PLS Partial Least Squares Calibrated Uncertainty Noise Estimation Orthogonality Constraints

2026-05-12abs ↗pdf ↗

Breaking the

T^{2/3}

Barrier for Sequential Calibrationcs.LG

Improved upper bound for online calibrated forecasting of binary sequences.

problem Online calibrated forecasting of binary sequences.

method Introducing a variant of Qiao & Valiant's sign preservation game called sign preservation with reuse (SPR) and proving its equivalence to calibrated forecasting.

result Improved upper bound of

O(T^{2/3 - \varepsilon})

for calibrated forecasting, improving the

O(T^{2/3})

bound of Foster & Vohra.

calibrated forecasting binary sequences sign preservation upper bound lower bound

2024-06-19abs ↗pdf ↗

Noise-Calibrated Inference from Differentially Private Sufficient Statistics in Exponential Familiescs.LG

This paper provides a method for noise-calibrated inference from DP synthetic data.

problem Inference from DP synthetic data is often miscalibrated and lacks principled uncertainty quantification.

method Release DP sufficient statistics, perform noise-calibrated likelihood-based inference, and optional synthetic data generation.

result Asymptotic normality and valid confidence intervals for the plug-in DP MLE.

differential privacy sufficient statistics likelihood-based inference noise calibration exponential families

2026-03-02abs ↗pdf ↗

Post-hoc Uncertainty Calibration for Domain Drift Scenarioscs.LG

Post-hoc calibration improves uncertainty under domain shift.

problem Improving uncertainty calibration under domain shift.

method Apply perturbations to validation set before post-hoc calibration.

result Perturbation step results in better calibration under domain shift.

uncertainty calibration domain shift post-hoc calibration deep learning

2020-12-20abs ↗pdf ↗

A comparison of linear and non-linear calibrations for speaker recognitionstat.ML

In recent work on both generative and discriminative score to log-likelihood-ratio calibration, it was shown that linear transforms give good accuracy only for a limited range of operating points. Moreover, these methods required tailoring of the calibration training objective functions in order to target the desired r…

2014-02-11abs ↗pdf ↗

Gaussian Process Pseudo-Likelihood Models for Sequence Labelingcs.LG

Several machine learning problems arising in natural language processing can be modeled as a sequence labeling problem. We provide Gaussian process models based on pseudo-likelihood approximation to perform sequence labeling. Gaussian processes (GPs) provide a Bayesian approach to learning in a kernel based framework. …

2014-12-25abs ↗pdf ↗

On the Discrepancy between Density Estimation and Sequence Generationcs.LG

This paper compares log-likelihood and BLEU scores for sequence generation tasks.

problem The discrepancy between density estimation and sequence generation performance.

method Comparing several density estimators on five machine translation tasks.

result The correlation between log-likelihood and BLEU varies depending on model families.

sequence generation machine translation log-likelihood BLEU density estimation

2020-02-17abs ↗pdf ↗

FRMDN: Flow-based Recurrent Mixture Density Networkcs.LG

Improves sequence modeling with a flow-based recurrent mixture density network.

problem Sequence modeling and sequence-to-sequence mapping applications.

method Generalized recurrent mixture density networks using normalized flow transformations.

result Significantly improved fit to image sequences measured by log-likelihood.

sequence modeling recurrent mixture density networks flow-based models log-likelihood image sequences

2020-08-05abs ↗pdf ↗

Amortized Conditional Normalized Maximum Likelihood: Reliable Out of Distribution Uncertainty Estimationcs.LG

ACNML method improves uncertainty estimation for deep networks.

problem Uncertainty estimation and calibration for deep neural networks under distribution shift.

method Approximate Bayesian inference to approximate CNML distribution.

result ACNML compares favorably to prior techniques for uncertainty estimation.

deep learning uncertainty estimation out-of-distribution approximate inference CNML

2020-11-05abs ↗pdf ↗

Modified Profile Likelihood Inference and Interval Forecast of the Burst of Financial Bubblesq-fin.ST

We present a detailed methodological study of the application of the modified profile likelihood method for the calibration of nonlinear financial models characterised by a large number of parameters. We apply the general approach to the Log-Periodic Power Law Singularity (LPPLS) model of financial bubbles. This model …

2016-02-26abs ↗pdf ↗

Calibrating Neural Simulation-Based Inference with Differentiable Coverage Probabilitystat.ML

New method calibrates neural SBI to avoid overconfident posteriors.

problem Overconfident posteriors in SBI due to inaccurate uncertainty quantification.

method Introduces a calibration term into neural model training objective, enabling end-to-end backpropagation.

result Achieves competitive or better coverage and posterior density than existing methods.

Bayesian inference Simulation-based inference Calibration Neural networks Coverage probability

2023-10-20abs ↗pdf ↗

Federated Martingale Posterior Sampingcs.LG

FMP sampling improves model calibration without sharing data.

problem Difficulties in specifying priors for modern neural networks.

method FMP sampling replaces prior and likelihood with a predictive distribution, running the sampler centrally.

result FMP sampling significantly improves calibration over baseline methods.

Federated Learning Bayesian Neural Networks Calibration Predictive Sampling

2026-05-18abs ↗pdf ↗

An Explicit Formula for Likelihood Function for Gaussian Vector Autoregressive Moving-Average Model Conditioned on Initial Observables with Application to Model Calibrationq-fin.ST

We derive an explicit formula for likelihood function for Gaussian VARMA model conditioned on initial observables where the moving-average (MA) coefficients are scalar. For fixed MA coefficients the likelihood function is optimized in the autoregressive variables $Φ$ 's by a closed form formula generalizing regression c…

2016-04-29abs ↗pdf ↗

On the Distance from Calibration in Sequential Predictioncs.LG

The paper studies the distance from calibration in sequential prediction, proving upper and lower bounds.

problem The challenge is to measure and minimize the deviation from perfect calibration in sequential binary prediction.

method The approach involves proving an

O(\sqrt{T})

upper bound and an

Ω(T^{1/3})

lower bound, using structural results and minimax arguments.

result An

O(\sqrt{T})

upper bound on the calibration distance is achieved, with an

Ω(T^{1/3})

lower bound showing the inherent difficulty.

calibration sequential prediction Lipschitz continuity upper bound lower bound

2024-02-12abs ↗pdf ↗

Conformal Calibration of Statistical Confidence Setsstat.ME

New methods improve confidence set calibration in complex models.

problem Challenges in maintaining confidence set coverage in complex models.

method TRUST and TRUST++ methods using simulated data for calibration.

result Methods achieve distribution-free conditional coverage and robust inference.

confidence sets statistical inference conformal prediction likelihood-free inference small sample size

2024-11-28abs ↗pdf ↗

Deep Linear Discriminant Analysis Revisitedstat.ML

DNLL loss improves deep LDA accuracy and consistency.

problem Pathological solutions in unconstrained Deep LDA.

method Introducing Discriminative Negative Log-Likelihood (DNLL) loss.

result Deep LDA trained with DNLL produces clean latent spaces and better calibrated probabilities.

Deep Learning Discriminative Learning Log-Likelihood Generative Models Calibration

2026-01-04abs ↗pdf ↗

Likelihood Ratio Confidence Sets for Sequential Decision Makingcs.LG

The paper proposes a method to construct confidence sets using likelihood ratios for sequential decision-making.

problem Constructing valid uncertainty estimates for unknown quantities in sequential decision-making.

method The method uses likelihood ratios to create any-time valid confidence sequences without specialized treatment for each application.

result The proposed confidence sets maintain the prescribed coverage in a model-agnostic manner and their size depends on the choice of estimator sequence.

likelihood ratios confidence sequences sequential decision-making uncertainty estimation online learning

2023-11-08abs ↗pdf ↗

Sequential Neural Likelihood: Fast Likelihood-free Inference with Autoregressive Flowsstat.ML

We present Sequential Neural Likelihood (SNL), a new method for Bayesian inference in simulator models, where the likelihood is intractable but simulating data from the model is possible. SNL trains an autoregressive flow on simulated data in order to learn a model of the likelihood in the region of high posterior dens…

2018-05-18abs ↗pdf ↗

Non-Parametric Calibration of Probabilistic Regressionstat.ML

The task of calibration is to retrospectively adjust the outputs from a machine learning model to provide better probability estimates on the target variable. While calibration has been investigated thoroughly in classification, it has not yet been well-established for regression tasks. This paper considers the problem…

2018-06-20abs ↗pdf ↗