Curve-Fitting Models: Linear, 4PL, 5PL, Cubic Spline
Choosing the right curve-fitting model is essential for accurate quantitation across linear assays, ELISA immunoassays, and complex dose-response experiments.
데이터분석
-
-
Statistical Metrics: R-squared, %CV, %RE Explained
R-squared, percent coefficient of variation, and percent relative error are the three core statistics for assessing calibration fit quality and measurement precision.
-
Spectral Preprocessing: Smoothing and Derivatives
Spectral preprocessing removes baseline drift and noise from raw spectra before chemometric modeling, improving the accuracy of NIR and Raman analysis.
-
Building a Calibration Curve
A calibration curve maps instrument response to known analyte concentrations so unknown samples can be accurately quantified.