Fig. 3From: The kernel-weighted local polynomial regression (KwLPR) approach: an efficient, novel tool for development of QSAR/QSAAR toxicity extrapolation modelsThe effects of changing the smoothing parameter values are illustrated with simulated data. It is straightforward to see that small bandwidth value (the dark green curve) corresponds to large variability and small bias, whereas small variability for the highest bandwidth value (purple curve) results in large biasBack to article page