In a multi-class classification problem, it is standard to model the output of a neural network as a categorical distribution conditioned on the inputs. The output must therefore be positive and sum to one, which is traditionally enforced by a softmax. This probabilistic mapping allows to use the maximum likelihood pri…
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.
168,738 papers · 148 categories
Trend · papers per month
48 results for “Log Loss”
Paper analyzes statistical properties of log-cosh loss function.
problem No statistical analysis of log-cosh loss function in literature.
method Presented statistical properties of log-cosh loss function, compared to Cauchy distribution, and examined various statistical procedures.
result Characterized statistical properties of log-cosh loss function, including distribution, likelihood function, and Fisher information.
SGD's escape rate depends on log loss barrier, not linear loss barrier.
problem Understanding the escape rate of SGD from local minima.
method Derived a stochastic differential equation (SDE) with additive noise from SGD's multiplicative noise property.
result The log loss barrier determines the escape rate of SGD, not the linear loss barrier.
Improved diffusion bridge sampling with rKL-LD loss.
problem Improving sampling from unnormalized distributions using diffusion bridges.
method Employing the rKL-LD loss instead of the Log Variance (LV) loss for diffusion bridges.
result rKL-LD consistently outperforms LV loss in diffusion bridges.
Separable losses are inconsistent for structured prediction models.
problem Inconsistency of separable losses in structured prediction models.
method Analysis of separable negative log-likelihood losses for structured prediction.
result Separable losses are not Bayes consistent and may not predict the most probable structure.
Burq-Gérard-Tzvetkov and Hu established estimates () for the restriction of eigenfunctions to submanifolds. The estimates are sharp, except for the log loss at the endpoint estimates for submanifolds of codimension 2. It has long been believed that the log loss at the endpoint can be remov…
PACMAN provides bounds for classification tasks considering accuracy vs. negative log-loss mismatch.
problem Mismatch between accuracy and negative log-loss in classification tasks.
method Point-wise PAC approach over generalization gap, using likelihood ratio and concentration inequalities.
result PACMAN provides point-wise PAC bounds for the generalization problem.
Generalized Huber Loss for Robust Learning and its Efficient Minimization for a Robust Statisticsstat.ML
Proposes a new Huber loss combining absolute and quadratic properties.
problem Improving robustness in learning models.
method Introduces a generalized Huber loss with a log-exp transform and provides an efficient minimization algorithm.
result Shows that the new loss function can be minimized efficiently.
New algorithm reduces prediction errors across various loss functions.
problem Online forecasting algorithms' inability to adapt to different loss functions.
method Design of a novel Follow-the-Perturbed-Leader (FTPL) algorithm with self-concordant noise.
result Simultaneously achieves regret for bounded proper losses and regret for bounded smooth proper losses.
Improved COCO algorithms with better constraint control.
problem Achieving small regret and constraint violation in online convex optimization.
method Simple projection-based algorithm leveraging self-contraction geometry.
result Exponential improvement in cumulative constraint violation for strongly convex losses.
We present -loss, , a tunable loss function for binary classification that bridges log-loss () and - loss (). We prove that -loss has an equivalent margin-based form and is classification-calibrated, two desirable properties for a good surrogate loss function for the ideal y…
We demonstrate that, in the classical non-stochastic regret minimization problem with decisions, gains and losses to be respectively maximized or minimized are fundamentally different. Indeed, by considering the additional sparsity assumption (at each stage, at most decisions incur a nonzero outcome), we derive…
In deep neural network, the cross-entropy loss function is commonly used for classification. Minimizing cross-entropy is equivalent to maximizing likelihood under assumptions of uniform feature and class distributions. It belongs to generative training criteria which does not directly discriminate correct class from co…
The paper improves sparse Gaussian processes by optimizing predictive loss.
problem Optimizing predictive loss in sparse Gaussian processes.
method Direct loss minimization (DLM) for log-loss and square loss, with product sampling (uPS) and biased Monte Carlo (bMC) for non-conjugate cases.
result DLM shows significant performance improvement in both log-loss and square loss cases.
Despite being the standard loss function to train multi-class neural networks, the log-softmax has two potential limitations. First, it involves computations that scale linearly with the number of output classes, which can restrict the size of problems we are able to tackle with current hardware. Second, it remains unc…
New method optimizes clustering with better log-likelihood landscape.
problem Nonconvex log-likelihood optimization in model-based clustering.
method Entropic optimal transport loss for Sinkhorn-EM algorithm.
result New loss function avoids spurious local optima.
Deep learning has become the method of choice in many application domains of machine learning in recent years, especially for multi-class classification tasks. The most common loss function used in this context is the cross-entropy loss, which reduces to the log loss in the typical case when there is a single correct r…
Paper studies multiclass classifiers from binary classifiers, proving methods and demonstrating advantages.
problem Constructing efficient multiclass classifiers from binary ones.
method Two methods: one vs. all and hierarchical classification, with a new leverage-hierarchical method introduced.
result Proves upper bounds and exact formulas for multiclass regret in terms of binary regrets.
Optimal sketching bounds for sparse linear regression under various loss functions are established.
problem Sparse linear regression under different loss functions.
method Distribution over oblivious sketches for sparse norm regression and hinge-like loss functions.
result Optimal sketching bounds with rows for sparse norm regression and rows for hinge-like loss functions.
We consider prediction with expert advice under the log-loss with the goal of deriving efficient and robust algorithms. We argue that existing algorithms such as exponentiated gradient, online gradient descent and online Newton step do not adequately satisfy both requirements. Our main contribution is an analysis of th…
In this work we study loss functions for learning and evaluating probability distributions over large discrete domains. Unlike classification or regression where a wide variety of loss functions are used, in the distribution learning and density estimation literature, very few losses outside the dominant ar…
New method improves transductive learning predictions with multiplicative oracle inequalities.
problem Improving transductive learning predictions with known covariates.
method Median of Level-Set Aggregation (MLSA) for transductive LOO prediction.
result Proved multiplicative oracle inequality for LOO error.
VarGrad reduces variance in ELBO gradient estimation for variational inference.
problem Improving the variance of gradient estimators in variational inference.
method VarGrad uses a new log-variance loss to estimate the ELBO gradient, achieving lower variance than the score function method.
result VarGrad offers a lower variance gradient estimator compared to other methods.
Log-Loss scores expose membership privacy breaches.
problem Privacy leakage from statistical aggregates like Log-Loss scores.
method Proved that Log-Loss scores enable full accuracy membership inference in a single query.
result Complete membership privacy breach is possible with Log-Loss scores.
DLM for BNNs fails to improve over ELBO optimization.
problem Performance of DLM for Bayesian Neural Networks (BNNs).
method Direct Loss Minimization (DLM) compared to ELBO optimization.
result DLM does not significantly improve over ELBO optimization for BNNs.
Solves learning halfspaces with Massart noise for log-concave distributions.
problem Learning halfspaces with Massart noise in distribution-specific PAC model.
method Identifies a smooth non-convex surrogate loss and uses SGD to solve the learning problem.
result First computationally efficient algorithm for learning halfspaces with Massart noise for a broad family of distributions.
Optimal convex loss function improves regression coefficient estimation.
problem Asymptotic variance improvement in linear regression estimation.
method Score matching extension for log-concave projection.
result Semiparametric estimator attains minimal asymptotic covariance.
The log-likelihood loss in heteroscedastic neural networks can lead to poor parameter estimates.
problem Capturing aleatoric uncertainty in deep learning models.
method Examine the log-likelihood loss in conjunction with gradient-based optimizers and propose an alternative formulation, -NLL.
result Using an appropriate largely mitigates the issue of poor parameter estimates.
We develop a variant of multiclass logistic regression that is significantly more robust to noise. The algorithm has one weight vector per class and the surrogate loss is a function of the linear activations (one per class). The surrogate loss of an example with linear activation vector and class has t…
Improved algorithm reduces communication rounds for distributed online learning.
problem Complicated constraints in distributed online learning with locally light computations.
method Proposed D-BOCG algorithm with delayed update mechanism and redefined surrogate loss function.
result Achieved regret bound with communication rounds for convex losses.
We introduce a tunable loss function called -loss, parameterized by , which interpolates between the exponential loss (), the log-loss (), and the 0-1 loss (), for the machine learning setting of classification. Theoretically, we illustrate a fundamental connection between $…
Study on sequential prediction with log-loss, focusing on well-specified and misspecified cases.
problem Sequential prediction with log-loss under different specification conditions.
method Analysis of cumulative regret in well-specified and misspecified cases for a Gaussian location hypothesis class.
result Cumulative regrets in well-specified and misspecified cases asymptotically coincide for the -dimensional Gaussian location hypothesis class.
This work broadens calibeating to various proper losses using Bregman divergence.
problem Calibration for a wide range of proper losses.
method Regret minimization and Bregman divergence approach.
result U-calibration results for a family of Tsallis losses with logarithmic regret and dimension independence.
This work generalizes calibeating for a broader range of proper losses using Bregman divergence.
problem Calibration for a wide range of proper losses beyond Brier and log loss.
method Regret minimization based on Bregman divergence for a family of proper losses.
result U-calibration results for a family of Tsallis losses with logarithmic regret and dimension independence.
We exhibit a strong link between frequentist PAC-Bayesian risk bounds and the Bayesian marginal likelihood. That is, for the negative log-likelihood loss function, we show that the minimization of PAC-Bayesian generalization risk bounds maximizes the Bayesian marginal likelihood. This provides an alternative explanatio…
Recent work on discriminative segmental models has shown that they can achieve competitive speech recognition performance, using features based on deep neural frame classifiers. However, segmental models can be more challenging to train than standard frame-based approaches. While some segmental models have been success…
Boosted decision trees typically yield good accuracy, precision, and ROC area. However, because the outputs from boosting are not well calibrated posterior probabilities, boosting yields poor squared error and cross-entropy. We empirically demonstrate why AdaBoost predicts distorted probabilities and examine three cali…
Prevalidated ridge regression simplifies logistic regression for high-dimensional data.
problem Efficient probabilistic classification in high-dimensional data with logistic regression.
method Developed a prevalidated ridge regression model that matches logistic regression's performance but is more computationally efficient.
result Prevalidated ridge regression achieves similar classification error and log-loss to logistic regression for high-dimensional data.
LOO prediction method improves generalization guarantees for arbitrary datasets.
problem Understanding LOO error guarantees in fully transductive settings for arbitrary datasets.
method Median of Level-Set Aggregation (MLSA) for empirical-risk level sets.
result Multiplicative oracle inequality for LOO error with complexity scaling.
Maximum likelihood training improves the performance of score-based diffusion models.
problem Training score-based diffusion models with maximum likelihood.
method Trained by minimizing a weighted combination of score matching losses, with a specific weighting scheme that bounds negative log-likelihood.
result Maximum likelihood training improves the log-likelihood of score-based diffusion models across multiple datasets.
Logsy detects anomalies in logs using a novel classification-based approach.
problem Anomaly detection in unstructured logs is challenging due to limited model generalization.
method Logsy learns log representations by distinguishing normal and anomaly logs using a classification-based approach with an attention-based encoder and hyperspherical loss function.
result Logsy improves anomaly detection performance by 0.25 in F1 score compared to previous methods.
We present new excess risk bounds for general unbounded loss functions including log loss and squared loss, where the distribution of the losses may be heavy-tailed. The bounds hold for general estimators, but they are optimized when applied to -generalized Bayesian, MDL, and empirical risk minimization estimators. …
Let $\cF$ be a set of classification procedures with values in . Given a loss function, we want to construct a procedure which mimics at the best possible rate the best procedure in $\cF$. This fastest rate is called optimal rate of aggregation. Considering a continuous scale of loss functions with various …
New DP algorithms achieve near-optimal regret bounds for online learning problems.
problem Online learning problems with zero-loss solutions and differential privacy constraints.
method Developed new Differentially Private algorithms with near-optimal regret bounds.
result Achieved near-optimal regret bounds for various online prediction and convex optimization problems.
Proposes a new generalization bound for Bayesian deep nets without strict assumptions.
problem Lack of generalization bounds for Bayesian deep nets without strict assumptions.
method Exploits contractivity of Log-Sobolev inequalities to add a loss-gradient norm term to the generalization bound.
result Introduces a new generalization bound for Bayesian deep nets that avoids strict assumptions.
Fast Rates in Stochastic Online Convex Optimization by Exploiting the Curvature of Feasible Setscs.LG
New algorithm exploits curvature of feasible sets for fast online convex optimization.
problem Online convex optimization with fast rates.
method Adapting FTL algorithm to curvature of feasible sets.
result Achieves logarithmic regret bound of in stochastic environments.
Proposes a new loss function for robust learning.
problem Creating a robust loss function for machine learning.
method Extended pseudo Huber loss with log-exp transform and logistic function.
result Linear convergence algorithm for minimizer finding.
New algorithms avoid a dominant lower-order term in heavy-tailed loss settings.
problem Prediction with heavy-tailed losses without prior knowledge.
method Adaptive algorithms that avoid the maximum of losses as a lower-order term in regret.
result Improved regret bounds of and .