Comparar métodos

	Previsão Conforme para Previsão de Séries Temporais	Gradient Boosting	Regressão por Mínimos Quadrados Ordinários (MQO)	Regressão Quantílica
Área	Econometria	Aprendizado de máquina	Econometria	Econometria
Família	Regression model	Machine learning	Regression model	Regression model
Ano de origem	2021	2001	2019	1978
Autor original	Angelopoulos & Bates (tutorial); Xu & Xie (time-series EnbPI)	Friedman, J. H.	Wooldridge (textbook treatment); classical least squares	Koenker & Bassett
Tipo	Distribution-free prediction interval wrapper	Ensemble (sequential boosting of decision trees)	Linear regression	Conditional quantile regression
Fonte seminal	Angelopoulos, A. N. & Bates, S. (2023). Conformal Prediction: A Gentle Introduction. Foundations and Trends in Machine Learning, 16(4), 494-591.	Friedman, J. H. (2001). Greedy Function Approximation: A Gradient Boosting Machine. Annals of Statistics, 29(5), 1189–1232.	Wooldridge, J. M. (2019). Introductory Econometrics: A Modern Approach (7th ed.). Cengage Learning. ISBN: 978-1337558860	Koenker, R. & Bassett, G., Jr. (1978). Regression Quantiles. Econometrica, 46(1), 33-50.
Outros nomes	conformal prediction, distribution-free prediction intervals, EnbPI, Konformal Tahmin (Conformal Prediction — Zaman Serisi)	Gradient Boosting (GBM), GBM, gradient boosted trees, gradient boosting machine	ordinary least squares, classical linear regression, linear regression, en küçük kareler regresyonu	conditional quantile regression, regression quantiles, Kantil Regresyon
Relacionados	4	5	5	5
Resumo	Conformal prediction is a distribution-free wrapper that turns any point forecaster — ARIMA, a neural network, or a machine-learning model — into valid prediction intervals using only its residuals. The time-series form was popularised by Xu & Xie (2021) and the modern tutorial treatment by Angelopoulos & Bates (2023).	Gradient Boosting is an ensemble learning method, formalised by Jerome H. Friedman in 2001, that combines a sequence of weak learners — typically shallow decision trees — so that each new tree is fitted to minimise the residual errors of the trees before it. It is the core algorithm behind popular implementations such as XGBoost, LightGBM and CatBoost.	Ordinary Least Squares is the classical linear regression method that explains a continuous outcome as a linear combination of predictors. It estimates the coefficients by minimising the sum of squared residuals, and under the Gauss-Markov assumptions these estimates are the best linear unbiased estimator (BLUE).	Quantile regression models conditional quantiles of an outcome - the median, the 25th or 75th percentile, and so on - rather than the conditional mean that OLS targets. Introduced by Koenker and Bassett in 1978, it reveals how predictors act across the whole distribution, including its tails.