classification of different cancerous animal tissues on the basis of their 1h nmr spectra data using different types of artificial neural networks

the main objective in classification of the nmr spectra of cancerous and healthy tissue , with high number of features is the prerequisites of the minimum number of samples. therefore the use of conventional classifier on this type of the data is not recommended. in the current work, different structures of the artificial neural networks (ann) were tried on classification of different cancerous and healthy tissues. the use of nonlinear classifiers such as ann could be a proper alternative in nmr spectra classification. the data consists of five type of cancerous and three types of nmr spectra of healthy animal tissues . in the case of multi layer perceptron (mlp), two-layer network with 11 hidden nodes gave the best solution on the current data. in addition mlp + single layer perceptron (slp) were tried on the current data but it did not make any improvement. finally the slp   network with log likelihood cost function was tried and in addition to giving the best classification it had a fast convergence time and gave a unique solution on the data independent of the initial seed value off the training. the slp classifier is better than any other classifiers for the current data.