Answers are grounded in this paper and cite their sources.
Compares LAsLS to traditional AsLS for baseline correction, indicating better accuracy for complex signals.
The asymmetric least squares (AsLS) method is commonly used and works well when baseline characteristics are relatively uniform, but can be less accurate for non‐uniform, complex, or noisy baselines. Here, we present local AsLS (LAsLS), a local extension of AsLS that preserves the original penalized least‐squares framework while allowing asymmetric weights and smoothing penalties to vary across user‐defined regions of the signal. This local parameterization provides more accurate baseline estimation for heterogeneous baselines. The method is versatile and can be applied to a variety of techniques, such as chromatography, Raman spectroscopy, and mid‐infrared spectroscopy. To test the method under controlled conditions, we compared LAsLS with traditional AsLS using simulated chromatogram‐like signals that mimic complex baseline scenarios. Across four cases with heterogeneous baselines, AsLS baseline rRMSE was 2.6–5.7 times higher than LAsLS, and peak quantification rRMSE was 1.4–3.5 times higher. In a controlled PLSR quantitation example, LAsLS also reduced external prediction rRMSE compared with AsLS (11.3% vs. 17.9%). We provide open‐source MATLAB implementations with a graphical interface to facilitate method adoption.
No takes yet. Share an insight, caveat, or question.
Gómez‐Sánchez et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: