An alternative risk evaluation model for cervical cancer was developed using the BP of an ANN. To test the capacity of this model to predict and classify unknown data, we applied the model to the test data using the coding rule listed in Table 2. The ANN model also demonstrated good classification ability. The predictive results on the test data showed that the accuracy in predicting cervical cancer was 80% in the patient group and 90% in the control group (Table 5), and the back substitution fitting had a high accuracy of classification. A sensitivity of 95.2% and a specificity of 99% were achieved with the ANN model (Table 6). This optimized model is compatible with input parameters.