A Neural Network Model to Classify Liver Cancer Patients Using Data Expansion and Compression

摘要

We develop a neural network model to classify liver cancer patients intohigh-risk and low-risk groups using genomic data. Our approach provides a noveltechnique to classify big data sets using neural network models. We preprocessthe data before training the neural network models. We first expand the datausing wavelet analysis. We then compress the wavelet coefficients by mappingthem onto a new scaled orthonormal coordinate system. Then the data is used totrain a neural network model that enables us to classify cancer patients intotwo different classes of high-risk and low-risk patients. We use theleave-one-out approach to build a neural network model. This neural networkmodel enables us to classify a patient using genomic data as a high-risk orlow-risk patient without any information about the survival time of thepatient. The results from genomic data analysis are compared with survival timeanalysis. It is shown that the expansion and compression of data using waveletanalysis and singular value decomposition (SVD) is essential to train theneural network model.