À suivreGuide suivant
Hypothèses de régression linéaire et analyse résiduelle
Technique
GUIDE Technique
Modèles linéaires généralisés
A generalized linear model (GLM) relates predictors to an outcome through a linear predictor, a response distribution and a link function.
Sur cette page4 minutes de lecture
Aperçu
This framework covers ordinary linear regression, logistic regression and Poisson regression by changing the outcome family and the mapping from the mean to the predictors.
Plongée profonde
A GLM has three connected pieces. The random component chooses a response distribution from an exponential-family model, such as Gaussian, binomial, Poisson or Gamma. The systematic component forms a linear predictor from the inputs, often written eta = X beta. The link function connects the expected response mu to that predictor through g(mu) = eta. The model remains linear in its coefficients on the link scale, even when the expected outcome itself is not a straight line in the predictors. With a Gaussian distribution and identity link, the conditional mean is the linear predictor, recovering the usual linear regression form. A binomial model with the logit link models log odds; applying the inverse logit gives a probability. A Poisson model with the log link models the logarithm of the expected count, so its predicted mean is positive. These links encode constraints and relationships suitable to different outcome types. For a hypothetical binary outcome, suppose a model's linear predictor is zero. The inverse logit maps zero to probability 0.5. If the linear predictor rises, the probability rises but remains below one. By contrast, an identity-link model could predict a binary response below zero or above one, which is why an ordinary Gaussian linear model may not suit that task. A GLM does not guarantee the selected family is correct; it makes the assumptions explicit so they can be assessed. The family determines a mean-variance relationship as well as support for the response. Poisson models commonly assume conditional variance equals the conditional mean, a condition that real count data may violate. A binomial model must reflect how trials and successes are represented. Link choices may be constrained by family and software implementation. Evaluate residuals or deviance diagnostics, uncertainty, calibration where relevant, and held-out performance. The GLM label does not mean that all outcome types share one error distribution or one interpretation of coefficients.
Impact stratégique
Coût et budget
Les décisions en matière d'architecture déterminent les performances et les coûts d'exploitation pendant des années.
Décisions plus claires
La formation technique aide les équipes à choisir la bonne pile, pas seulement la plus récente.
Contrôle qualité
De meilleurs choix d’ingénierie réduisent les incidents de fiabilité en production.
The Future of Generalized Linear Models
GLM workflows can become clearer when reports show the family, link, response units and coefficient interpretation together. Analysts can compare plausible families on data that match the intended population and decision, while checking calibration or residual patterns suited to the outcome. When count variance exceeds a Poisson model's assumptions, a negative-binomial or other model may deserve investigation. As applications evolve, documentation should preserve offsets, weights and preprocessing alongside the fitted coefficients. Better tooling can aid those checks, but a family choice remains a modeling decision grounded in how the observations were generated.
Mise en œuvre dans le monde réel
A housing analyst uses a Gaussian family with identity link to model average sale price; the predicted conditional mean is the linear predictor itself.
A hypothetical service team models whether a case is resolved in one day with a binomial family and logit link. The linear predictor describes log odds, which the inverse link maps to a probability between zero and one.
A transit planner predicts nonnegative expected incident counts with a Poisson family and log link. Exponentiating the linear predictor gives a positive expected count, while exposure time can be represented through an offset when appropriate.
A researcher models positive, right-skewed costs using a Gamma family and a suitable link. They check whether that response family and variance structure match the measurement process rather than choosing a GLM solely because the target is positive.
Risques et garde-fous
L’optimisation d’un benchmark peut masquer des faiblesses plus larges du système.
Les coûts d’infrastructure et de maintenance sont souvent sous-estimés.
Les lacunes en matière de sécurité et d’observabilité peuvent se creuser à mesure que les systèmes deviennent plus complexes.
Feuille de route de mise en œuvre
Définissez les objectifs de latence, de qualité et de coût avant la mise en œuvre.
Benchmark dans des conditions de charge et de données réalistes.
Surveillance des instruments pour détecter les erreurs, la dérive et l'impact sur l'utilisateur.
Préparez les chemins de restauration et de réponse aux incidents avant la mise à l’échelle.
Continuez à explorer
Free newsletter
Get the daily AI briefing
Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Take the Generalized Linear Models quiz
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
Questions fréquemment posées
What is Generalized Linear Models?
A generalized linear model (GLM) relates predictors to an outcome through a linear predictor, a response distribution and a link function. This framework covers ordinary linear regression, logistic regression and Poisson regression by changing the outcome family and the mapping from the mean to the predictors.
Which three elements define the basic GLM structure described here?
The GLM combines a response family, a predictor expression and a link connecting the mean to that predictor.
In a Gaussian GLM with identity link, how does the conditional mean relate to the predictor?
The identity link leaves the mean on the same scale as the linear predictor.
A binomial GLM uses a logit link. What scale does its linear predictor describe?
The logit link maps a probability to log odds; the inverse link maps back to probability.
Why can a Poisson GLM with a log link produce positive expected counts?
The inverse of the log link is the exponential function, whose values are positive.
What does the response family specify beyond the outcome's allowable values?
The family determines distributional behavior including the relationship between conditional mean and variance.
Continuez à apprendre
Guides associés
Plus de guides sélectionnés pour ce sujet