Locally Weighted Regression: An Approach to Regression Analysis by Local Fitting

By William S. Cleveland, Susan J. Devlin. 1988

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Local Linear Forests (2018, methodology): “Early versions of confidence intervals for random forests, backed by heuristic arguments and empirical evidence, were proposed by Sexton and Laake [2009] and Wager, Hastie, and Efron [2014]. Mentch and Hooker [2016] then established asymptotic normality of random forests where each tree depends on a”
Forecasting at Scale (2018, methodology): “In practice, we use a local regression (Cleveland & Devlin 1988) or isotonic regression (Dykstra 1981) as ﬂexible non-parametric models of error curves.”
Reinforcement Learning: A Survey (1996, methodology): “The juggling robot learned a world model from experience, which was generalized to unvisited states by a function approximation scheme known as locally weighted regression (Cleveland & Delvin, 1988; Moore & Atkeson, 1992).”
Supervised Weighting-Online Learning Algorithm for Short-Term Traffic Flow Prediction (2013, methodology): “Syst., vol. 13, no. 2, pp. 644–654, Jun. 2012.”
Heterogeneous expectations, exchange rate dynamics and predictability (2007, unknown): “We use the locally weighted regression (LWR) framework of Cleveland and Devlin (1988). Previous applications of LWR to exchange rates are Diebold and Nason (1990) and Meese and Rose (1991). We assume the following nonparametric regression model for exchange rate returns rt+1 = m(Xt) + ǫt, (11)”
On birth and death in the sea (2002, background): “Although Loess fits (Cleveland & Devlin, 1988) suggested that these relationships may not be linear in log-log space, nonlinear curves fitted to these data showed virtually no improvement in r2 and thus we present the linear relations here to facilitate comparisons.”
Teachers’ Knowledge About Early Reading: Effects on Students’ Gains in Reading Achievement (2011, methodology): “As a post hoc analysis, we examined this issue, using locally weighted polynomial regression analysis (eg, Cleveland & Devlin, 1988 ).”
Spatio‐temporal analysis and modeling of short‐term wind power forecast errors (2011, methodology): “For an extensive description of conditional parametric models, we refer to [24, 25].”
Forecasting Electricity Spot Prices Accounting for Wind Power Predictions (2013, methodology): “An excellent general description of the methodology and estimation procedure used to describe the spot prices as a function of wind power and load forecasts is given in [24] (without recursivity) and in [25], [26](including recursi vity and robustness respectively).”
The Impact of Mixed Land Use on Residential Property Values (2012, unknown): “Ekeland et al. (2004) then also 11Locally weighted regression is developed by Cleveland and Devlin (1988) and was first applied in urban economics by Meese and Wallace (1991).”
Estimation And Hypothesis Testing For Nonparametric Hedonic House Price Functions (2010, unknown): “Urban applications of the LWR estimator are derived from the seminal paper by Cleveland and Devlin (1988).”
Cardiomyocyte proliferation contributes to heart growth in young humans (2013, background): “5F) and fitted the data with LOWESS, validating the fits using generalized cross-validation.”
Geographically weighted regression bandwidth selection and spatial autocorrelation: an empirical example using Chinese agriculture data (2010, methodology): “Local spatial regression techniques were originally introduced by Cleveland and Devlin ( 1988 ).”
From Predictive to Prescriptive Analytics (2014, methodology): “A generalization of KR is local polynomial regression (Cameron and Trivedi 2005, p. 311), of which the LOESS method of Cleveland and Devlin (1988) is a specific case.”
Potential impacts of climatic change upon geographical distributions of birds (2006, methodology): “We have ﬁtted these surfaces using locally weighted regression (Cleveland & Devlin 1988) because this makes no assumptions about the form of the relationship between a species’ probability of occurrence and the bioclimatic variables.”