PERFORMANCE OF MULTILAYER FEEDFORWARD NEURAL NETWORKS TO PREDICT LIVER-TRANSPLANTATION OUTCOME

A novel multisotutional clustering and quantization (MCQ) algorithm ha s been developed that provides a flexible way to preprocess data. It w as tested whether it would impact the neural network's performance fav orably and whether the employment of the proposed algorithm would enab le neural networks to handle missing data. This was assessed by compar ing the performance of neural networks using a well-documented data se t to predict outcome following liver transplantation. This new approac h to data preprocessing leads to a statistically significant improveme nt in network performance when compared to simple linear scaling. The obtained results also showed that coding missing data as zeroes in com bination with the MCQ algorithm, leads to a significant improvement in neural network performance on a data set containing missing values in 59.4% of cases when compared to replacement of missing values with ei ther series means or medians.