Abstract: The pulp kappa number is determined quantitatively by measurement of absorbance of cooking liquor. Three methods, artificial neural networks (ANN), stepwise regression analysis-artificial neural networks (SRA-ANN) and principal component regression-artificial neural networks (PCA-ANN), are applied to data processing. The results of data processing show that the predicting models determined by three methods all have good learning accuracy and predicting accuracy, and the best model of all is determined by principal component regression-artificial neural networks.

Key words: Artificial neural networks, Quantitative analysis, Kappa number