Introduction: Primary liver cancer (­HCC), is the fifth most common type of cancer and the third leading cause of death in the world. Symptoms of liver cancer will progress rapidly after the onset of the disease, and unfortunately, the patients' survival rate is very low. One of the main problems for gastroenterologists is the prediction and early detection of liver cancer. Data mining techniques can be used to understand and predict cancer. The aim of this study was to identify the best model based on intelligent data mining to predict and diagnose liver cancer in an early stage.

Method: In the present article, a retrospective study was conducted on 516 cases of primary and secondary liver cancer, and 22 risk factors were examined. Data were collected from the patients' files and analyzed using 5 data mining models including VFI Classifier, Regression Classifier, Hyper Pipes Classifier, Functional trees with logistic regression, and Meta Multi Class Classifier with the highest precision (Precision). These models were compared.

Results: The precision, sensitivity, specificity, and the area under the curve of VFI Classifier model were respectively 71.29%, 49%, 50%, and 63.31%, and VFI Classifier model is the best model based on intelligent data mining to predict and diagnose liver cancer in an early stage.

Conclusion: If properly designed, data mining model VFI Classifier can predict liver cancer or detect it in an early stage.