Title :
The DAGs-MLP Structure to the Efficiency of Neural Network Classification for Diagnosis of Hepatobiliary Disoders
Author :
Niyom, Anan ; Chiewchanwattana, Sirapat ; Sunat, Khamron ; Lursinsap, Chidchanok
Author_Institution :
Dept. of Comput. Sci., Khon Kaen Univ.
Abstract :
We present the effect of directed acyclic graph (DAGs) to the efficiency of neural network models in a diagnosis of hepatobiliary disorder. We compare the model of DAGs structure in the general neural network model with the general neural network model. We name these new algorithms as DAGs-MLP, DAGs-SVM, DAGs-RBF and DAGs-RPI. The efficiency of each algorithm from the two models is determined from the accuracy of data classification. The results show that DAG methods can improve the efficiency of general algorithms focused in this research. DAGs-MLP technique has the best efficiency
Keywords :
directed graphs; medical diagnostic computing; multilayer perceptrons; pattern classification; radial basis function networks; support vector machines; data classification; directed acyclic graph; hepatobiliary disorder; medical diagnosis computing; multilayer perceptron; neural network; radial basis function network; support vector machine; Computer science; Data analysis; Electronic mail; Liver; Medical diagnostic imaging; Multilayer perceptrons; Neural networks; Radial basis function networks; Support vector machine classification; Support vector machines; Directed Acyclic Graph; Multi-Layer Perceptrons; Radial Basis Function Network; Random & Pseudoinverse; Support Vector Machines;
Conference_Titel :
SICE-ICASE, 2006. International Joint Conference
Conference_Location :
Busan
Print_ISBN :
89-950038-4-7
Electronic_ISBN :
89-950038-5-5
DOI :
10.1109/SICE.2006.315306
