Employing Linear Regression in Regression Tree Leaves

By Aram Karalic. 1992

How researchers see this paper

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

Local Linear Forests (2018, methodology): “Karalič (1992) suggested fitting a local linear regression in each leaf, and Torgo (1997) highlighted the performance of kernel methods in general for MOB tree methods.”
Locally Weighted Learning (1997, methodology): “Kozek (1992) describe using LWR to model automobile emissions. Walden and Prescott (1983) use LWR to remove trends in time series involving climate data. Solow (1988) estimated the variance or noise level in time series climate data after having removed the mean using LWR.”
Functional Trees (2001, unknown): “Also Karalic (1992) studied the influence of using linear regression in the leaves of a regression tree.”
Towards global empirical upscaling of FLUXNET eddy covariance observations: validation of a model tree ensemble approach using a biosphere model (2009, background): “Splits along continuous variables are determined by ﬁnding the split location l i of each continuous explanatory variable X where the joint sum of squared errors of the left (X j < l i ) and right child (X j > =l i ) is minimal (cf. Karalic, 1992): min (SSE left +SSE right ) .”
Learning model trees from evolving data streams (2010, methodology): “Here fall RETIS ( Karalic 1992 ), SUPPORT (Chaudhuri et al. 1994), SECRET (Dobra and Gherke 2002), GUIDE (Loh 2002), SMOTI (Malerba et al. 2002) and LLRT (Vogel et al. 2007).”