Tiếp theoHướng dẫn tiếp theo
Giả định hồi quy tuyến tính và phân tích phần dư
kỹ thuật
HƯỚNG DẪN KỸ THUẬT
Mô hình tuyến tính tổng quát (GLM) liên kết các yếu tố dự đoán với kết quả thông qua yếu tố dự đoán tuyến tính, phân phối phản hồi và hàm liên kết.
Nghe hướng dẫn nàyAI thuật lại · ~3 min
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.
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.
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.
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.
Bản tin miễn phí
Ba câu chuyện AI đã được xác minh vào mỗi buổi sáng các ngày trong tuần, được viết bằng tiếng Anh đơn giản. Miễn phí mãi mãi, không có quảng cáo.
Một email mỗi ngày trong tuần. Hủy đăng ký trong một cú nhấp chuột. Chúng tôi không bao giờ bán hoặc chia sẻ địa chỉ của bạn.
Tự kiểm tra
Phản hồi tức thì về mọi câu trả lời và chứng chỉ có thể chia sẻ với ID có thể xác minh sau khi bạn vượt qua khóa học.
Hỗ trợ giáo dục AI miễn phí. AI Understanding là tổ chức phi lợi nhuận 501(c)(3) — không bao giờ có quảng cáo, không có tường phí. Hãy quyên góp
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.
The GLM combines a response family, a predictor expression and a link connecting the mean to that predictor.
The identity link leaves the mean on the same scale as the linear predictor.
The logit link maps a probability to log odds; the inverse link maps back to probability.
The inverse of the log link is the exponential function, whose values are positive.
The family determines distributional behavior including the relationship between conditional mean and variance.
Tiếp tục học hỏi
Đã chọn thêm hướng dẫn cho chủ đề này
Tiếp theoHướng dẫn tiếp theo
Giả định hồi quy tuyến tính và phân tích phần dư
kỹ thuật