In mathematical optimization and decision theory, a loss function or cost function (sometimes also called an error function) is a function that maps an event or values of one or more variables onto a real number intuitively representing some "cost" associated with the event. An optimization problem seeks to minimize a loss function. An objective function is either a loss function or its opposite (in specific domains, variously called a reward function, a profit function, a utility function, a fitness function, etc.), in which case it is to be maximized. The loss function could include terms from several levels of the hierarchy.
In statistics, typically a loss function is used for parameter estimation, and the event in question is some function of the difference between estimated and true values for an instance of data. The concept, as old as Laplace, was reintroduced in statistics by Abraham Wald in the middle of the 20th century. In the context of economics, for example, this is usually economic cost or regret. In classification, it is the penalty for an incorrect classification of an example. In actuarial science, it is used in an insurance context to model benefits paid over premiums, particularly since the works of Harald Cramér in the 1920s. In optimal control, the loss is the penalty for failing to achieve a desired value. In financial risk management, the function is mapped to a monetary loss.
Regret (decision theory)
Leonard J. Savage argued that using non-Bayesian methods such as minimax, the loss function should be based on the idea of regret, i.e., the loss associated with a decision should be the difference between the consequences of the best decision that could have been made had the underlying circumstances been known and the decision that was in fact taken before they were known.
The use of a quadratic loss function is common, for example when using least squares techniques. It is often more mathematically tractable than other loss functions because of the properties of variances, as well as being symmetric: an error above the target causes the same loss as the same magnitude of error below the target.
Cette page est générée automatiquement et peut contenir des informations qui ne sont pas correctes, complètes, à jour ou pertinentes par rapport à votre recherche. Il en va de même pour toutes les autres pages de ce site. Veillez à vérifier les informations auprès des sources officielles de l'EPFL.
En statistique ou en épidémiologie, un biais est une démarche ou un procédé qui engendre des erreurs dans les résultats d'une étude. Formellement, le biais de l'estimateur d'un paramètre est la différence entre la valeur de l'espérance de cet estimateur (qui est une variable aléatoire) et la valeur qu'il est censé estimer (définie et fixe). biais effet-centre biais de vérification (work-up biais) biais d'autosélection, estimé à 27 % des travaux d'écologie entre 1960 et 1984 par le professeur de biologie américain Stuart H.
En statistique, la théorie de l'estimation s'intéresse à l'estimation de paramètres à partir de données empiriques mesurées ayant une composante aléatoire. Les paramètres décrivent un phénomène physique sous-jacent tel que sa valeur affecte la distribution des données mesurées. Un estimateur essaie d'approcher les paramètres inconnus à partir des mesures.
La théorie de la commande optimale permet de déterminer la commande d'un système qui minimise (ou maximise) un critère de performance, éventuellement sous des contraintes pouvant porter sur la commande ou sur l'état du système. Cette théorie est une généralisation du calcul des variations. Elle comporte deux volets : le principe du maximum (ou du minimum, suivant la manière dont on définit l'hamiltonien) dû à Lev Pontriaguine et à ses collaborateurs de l'institut de mathématiques Steklov , et l'équation de Hamilton-Jacobi-Bellman, généralisation de l'équation de Hamilton-Jacobi, et conséquence directe de la programmation dynamique initiée aux États-Unis par Richard Bellman.
The course introduces the theoretical foundations to choice modeling and describes the steps of operational modeling.
Discrete choice models are used extensively in many disciplines where it is important to predict human behavior at a disaggregate level. This course is a follow up of the online course “Introduction t
Discrete choice models are used extensively in many disciplines where it is important to predict human behavior at a disaggregate level. This course is a follow up of the online course “Introduction t
Machine learning and data analysis are becoming increasingly central in sciences including physics. In this course, fundamental principles and methods of machine learning will be introduced and practi
This course aims to introduce the basic principles of machine learning in the context of the digital humanities. We will cover both supervised and unsupervised learning techniques, and study and imple
This course provides an overview of key advances in continuous optimization and statistical analysis for machine learning. We review recent learning formulations and models as well as their guarantees
As large, data-driven artificial intelligence models become ubiquitous, guaranteeing high data quality is imperative for constructing models. Crowdsourcing, community sensing, and data filtering have long been the standard approaches to guaranteeing or imp ...
EPFL2024
In this thesis we study stability from several viewpoints. After covering the practical importance, the rich history and the ever-growing list of manifestations of stability, we study the following. (i) (Statistical identification of stable dynamical syste ...
EPFL2024
, , ,
Catastrophic overfitting (CO) in single-step adversarial training (AT) results in abrupt drops in the adversarial test accuracy (even down to 0%). For models trained with multi-step AT, it has been observed that the loss function behaves locally linearly w ...