Function approximation using artificial neural networks
MATH'07 Proceedings of the 12th WSEAS International Conference on Applied Mathematics
An algorithm for saving the memory utilization in the 1-D cerebellar model controller
NN'05 Proceedings of the 6th WSEAS international conference on Neural networks
Knowledge induction from medical databases with higher-order programming
WSEAS Transactions on Information Science and Applications
| Hi-index | 0.00 |
In this paper, a preliminary result about the melancholia diagnosis scheme based on the analyses of depression questionnaire and the meridian energy of human body using CMAC (Cerebellar Model Articulation Controller) neural network approach is proposed. Firstly, a large amount data obtained from hospital, recorded the aged patients' depression rating scales and the 12 sets meridian energy signals, are sieved out three disease groups' patterns. Assuming the recorded data can describe the necessary features of the melancholia patient. Then, we built a CMAC neural network to learn the melancholia features depending on the three disease groups' patterns. By the sufficient training, the diagnosis architecture will memorize the features of the selected melancholia patient patterns. Finally, the built diagnosis system can used to diagnose the depression scale by inputting the 12 sets meridian energy signals of human body into CMAC neural network. To benefit the pattern collection, re-training, diagnosis and the data analyses, a PC-based friendship operation interface is developed in this paper also. Such as the function of new pattern addition, retraining, and the memory weights distribution plots are appeared in the interface.